- Email: [email protected]
UAV Lab: a multidisciplinary UAV design course
21st IFAC Symposium on Automatic Control in Aerospace 21st 21st IFAC IFAC Symposium Symposium on on Automatic Automatic Control Control in in Aerospace Aerospace August 27-30, 2019. UK 21st IFAC Symposium on Automatic Control in Aerospace August 27-30, 2019. Cranfield, Cranfield, UK Available online at www.sciencedirect.com August 27-30, 2019. Cranfield, UK Control in Aerospace 21st IFAC Symposium on Automatic August 27-30, 2019. Cranfield, UK
ScienceDirect
IFAC PapersOnLine 52-12 (2019) 490–495
UAV UAV UAV UAV
Lab: Lab: Lab: Lab:
a a a a
multidisciplinary UAV design multidisciplinary multidisciplinary UAV UAV design design course multidisciplinary UAV design course course ∗ ∗ Mattia Giurato ∗ Paolo Gattazzo ∗∗∗ Marco Lovera ∗∗∗
Mattia Mattia Giurato Giurato ∗ Paolo Paolo Gattazzo Gattazzo ∗ Marco Marco Lovera Lovera ∗ Mattia Giurato ∗ Paolo Gattazzo ∗ Marco Lovera ∗ ∗ ∗ ∗ ∗ Mattia Giurato Paolo Gattazzo Marco Lovera ∗ Dipartimento ∗ di Scienze Tecnologie Aerospaziali, Politecnico di Dipartimento di di Scienze Scienze eee Tecnologie Tecnologie Aerospaziali, Aerospaziali, Politecnico Politecnico di di ∗ Dipartimento ∗ Dipartimento di Scienze e 34, Tecnologie Aerospaziali, Politecnico di Milano, Via La Masa 20156, Milano, Italy (e-mail: Milano, Via La Masa 34, 20156, Milano, Italy (e-mail: Milano, Via La Masa 34, 20156, Milano, Italy (e-mail: ∗ Milano, Via La Masa 34, 20156, Milano, Italy (e-mail: Dipartimento di Scienze e Tecnologie Aerospaziali, Politecnico di [email protected]). [email protected]). [email protected]). Milano, Via [email protected]). Masa 34, 20156, Milano, Italy (e-mail: [email protected]). Abstract: Aerospace Aerospace control control education education can can significantly significantly benefit benefit from from actual actual hands-on hands-on experiexperiAbstract: Abstract: Aerospace control education can significantly benefit fromto actual hands-on experience. In most cases, however, such experience can only be provided to students in small-scale ence. In most cases, however, such experience can only be provided students in small-scale ence. In most cases, however, such experience can only bebenefit provided students in small-scale Abstract: Aerospace control can significantly fromto actual hands-on experience. Inactivities. most cases, however, such experience provided to students in small-scale project activities. In this this papereducation the experience experience ofcan theonly UAVbe Lab course, held at Politecnico Politecnico for the the project In paper the of the UAV Lab course, held at for ence. Inactivities. most cases, however, such experience can only be provided to students in small-scale project In this paper the experience of the UAV Lab course, held at Politecnico for the first time in the fall of 2018, is presented and discussed. The course, aimed at an interdisciplinary first time in of 2018, is presented and discussed. The course, aimed at an interdisciplinary first time in the the fall fallIn ofthis 2018, presented and discussed. TheLab course, aimed an interdisciplinary project paper the experience offor theaa UAV course, heldatat Politecnico for the group ofactivities. students, covers theiswhole whole design cycle for multirotor UAV, from conceptual design to group of students, covers the whole design cycle multirotor UAV, from conceptual design to group of students, covers the design cycle for a multirotor UAV, from conceptual design to first time in the fall of 2018, is presented and discussed. The course, aimed at an interdisciplinary group of students, covers the whole design cycle for a multirotor UAV, from conceptual design to in-flight validation, with specific reference to modelling, simulation, identification and control. in-flight validation, with specific reference to modelling, simulation, identification and control. in-flight with the specific reference to modelling, simulation, identification anddesign control. group of validation, students, covers whole cycle for a multirotor UAV, from to in-flight validation, with specific reference to modelling, simulation, The course course has been been conceived asdesign an extra-curricular extra-curricular activity, soidentification the conceptual emphasisand is control. not on on The has conceived as an activity, so the emphasis is not in-flight validation, with specific reference to experience modelling, in simulation, identification and control. The course has been conceived as an extra-curricular activity, so the emphasis is not on conventional lectures but rather on hands-on hardware/software integration, data conventional lectures but rather on experience in hardware/software integration, data conventional lectures rather onashands-on hands-on experience inactivity, hardware/software integration, The course beenbut conceived an The extra-curricular so syllabus the emphasis is notdata on collection andhas analysis and flight testing. testing. The paper presents the course course syllabus and organisation organisation collection and analysis and flight testing. paper presents the course and organisation collection and analysis and flight The paper presents the syllabus and conventional lectures but rather on hands-on experience in hardware/software integration, data collection and analysis and flight testing. The paper presents the course syllabus and organisation and provides provides an an overview overview of of the the obtained obtained results results and and the the feedback feedback provided provided by by the the students. students. and provides an overview of the obtained results and the feedback provided by the students. and collection andan analysis andof flight paperand presents the course syllabus and provides overview the testing. obtainedThe results the feedback provided byand theorganisation students. Copyright © 2019. Authors.of Published by Elsevier Ltd. and All rights reserved. provided by the students. and provides anThe overview the obtained results the feedback 1. INTRODUCTION the 1. INTRODUCTION INTRODUCTION the corresponding syllabus is discussed in Section 3. Sub1. the corresponding corresponding syllabus syllabus is is discussed discussed in in Section Section 3. 3. SubSub1. INTRODUCTION the corresponding syllabus is discussed in Section 3.teams Subsequently, the sizing approach used by the student sequently, the sizing approach used by the student teams sequently, the sizing approach used by the student teams 1. INTRODUCTION the corresponding syllabus is discussed in Section 3.teams Subsequently, the approach usedactivities by the student to carry carry out thesizing conceptual design activities and the the three out the conceptual design and three In recent recent years years Unmanned Unmanned Aerial Aerial Vehicles Vehicles (UAVs), (UAVs), which which to sequently, the sizing approach usedactivities by the student student teams to carry out the conceptual design and the three In requirement specifications provided to the teams specifications provided to the student teams In recent years Aerialonly Vehicles (UAVs), which requirement in the past had been for applicarequirement specifications provided to an theoverview student to carry out the conceptual design activities and theteams three in the the past had Unmanned been of interest only for military military applicaare then illustrated in Section 4, while of in past had been of of interest interest only for military applicaare then illustrated in Section 4, while an overview of the are then illustrated in Section 4, while an overview of the the In recent years Unmanned Aerial Vehicles (UAVs), which in the past had been of interest only for military applications, have have started started to to play play aa significant significant role role in in civil civil applica- requirement specifications provided to the student teams are then illustrated in Section 4, while an overview of the tions, students’ design activities and an example of the obtained students’ design activities and an example of the obtained students’ design activities and an example of the obtained in theas past had been of interest onlyand for military tions, have started to play a significant role in civil applications well, ranging from personal commercial use to are then(referring illustrated in Section 4,one while ofmulthe students’ design activities andto example ofdesigned the obtained tions as well, ranging from and commercial use to results specifically toan of an theoverview designed specifically one of the multions as well, ranging fromapersonal personal commercial to results tions, have started toapplications. play significant in civil applicacountless industrial In and therole framework ofuse civil results (referring (referring specifically of the mulstudents’ design activities andtoanone example ofdesigned the 5obtained countless industrial applications. In the framework of civil tirotors) are provided, respectively, in Section and in are provided, respectively, in Section 55 and in countless industrial In and the framework ofuse civil tions as well, rangingapplications. from personal commercial to tirotors) applications, multirotor UAVs represent the most common tirotors)(referring are provided, respectively, inaathe Section and in results specifically to one of designed mulapplications, multirotor UAVs represent the most most common common Section 6. Finally Section 7 provides discussion of the applications, multirotor UAVs represent the Section 6. Finally Section 7 provides discussion of the Finally Section 7 providesina Section discussion of the countless industrial In the framework of As civil Section 6.are applications, multirotor UAVs represent the most common architecture, due to toapplications. their versatility and reliability. provided, respectively, 5 and in Section 6. Finally Section 7 provides a discussion of the architecture, due their versatility and reliability. As aa tirotors) overall course course experience, experience, some some lessons lessons learned learned and and aa few few applications, UAVs represent most common architecture, due to their versatility andthe reliability. As of consequence, multirotor education activities related to the design design ofa overall Section 6. Finally Section 7 provides discussion the overall course experience, some lessonsalearned and of a few consequence, education activities related to the perspectives for further developments. perspectives for further developments. consequence, education activities related to the design of architecture, due have to their versatility andmore reliability. As a overall multirotor UAVs UAVs become more and and widespread, perspectives further developments. coursefor experience, some lessons learned and a few multirotor have become more more widespread, multirotor UAVs have both become more and more widespread, consequence, education activities related to the design of of perspectives for further developments. with courses covering specific disciplinary aspects with courses courses covering both specific disciplinary aspects of with covering both specific disciplinary aspects of 2. APPROACH APPROACH 2. multirotor UAVs have become more and more widespread, with courses covering both specific disciplinary aspects of their design design and and operation operation (aeromechanics, (aeromechanics, power power electronelectrontheir 2. APPROACH with courses covering both(aeromechanics, specific disciplinary of their design and operation poweraspects electronics, hardware hardware and software, navigation, control, teleme2. APPROACH ics, and software, navigation, control, telemeThe The course was organised and managed according to the ics, hardware software, navigation, control, telemetheir design andand operation (aeromechanics, power electronThe course course was was organised organised and and managed managed according according to to the the try/communications etc.) and system-level design issues try/communications etc.) and system-level design issues The course was organised and managed according to the following approach: following approach: ics, hardware and software, navigation, control, telemetry/communications etc.) and system-level design issues following approach: (see, e.g., Gaponov and Razinkova (2012) Khan et al. (see, e.g., Gaponov and Razinkova (2012) Khan et al. course was organised and managed according to the (see, e.g., Gaponov and Razinkova (2012) Khan issues et al. The following approach: try/communications and system-level (see, e.g., Gaponov etc.) and Razinkova (2012) design Khan et al. following (2017)). ••• aaa call for call for the definition of three interdisciplinary (2017)). approach: call for the the definition definition of of three three interdisciplinary interdisciplinary (see, e.g., Gaponov and Razinkova (2012) Khan et al. (2017)). • ateams call was for the definition ofstudents three interdisciplinary sent to Master in teams was sent to Master students in Aeronautical teams was sent to Master students in Aeronautical Aeronautical In this paper the experience of the UAV Lab course, In this paper the experience of the UAV Lab course, (2017)). • ateams call for the definition of three interdisciplinary In this paper the experience of the UAV Lab course, was sent to Master students in Aeronautical Engineering, Space Engineering, Computer Science Engineering, Space Engineering, Computer Science In this paper the experience of the UAV Lab course, Engineering, Space Engineering, Computer Science held at Politecnico di Milano for the first time in the Engineering, Space Engineering, Computer Science held at Politecnico di Milano for the first time in the teams was sent to Master students in Aeronautical and Engineering, Automation and Control Engineerand Engineering, Automation and Control EngineerIn this paper the experience of the UAV Lab course, held at Politecnico di Milano for the first time in the and Engineering, Automation and Control Engineerfall of 2018, is presented and discussed. The course aims fall of is presented and The course aims Engineering, Space Engineering, Computer Science and Engineering, Automation and Control Engineering. four programs were selected among ing. These four Master programs were selected among fall ofat2018, 2018, isteams presented and discussed. discussed. The time course aims held Politecnico di students Milano for first the ing. These These four Master Master programs were selected among at providing of the opportunity to carry at providing providing teams of students the the opportunity toincarry carry and Engineering, Automation and Control Engineerat teams of students the opportunity to ing. These four Master programs were selected among the ones offered at Politecnico di Milano in view of the ones offered at Politecnico di Milano in view of fall of 2018, is presented and discussed. The course aims at providing teams of students the opportunity to carry the ones offered at Politecnico di Milano in view of out design activities in the field of multirotor UAVs. More out design activities in the field of multirotor UAVs. More ing. These four Master programs were selected among out design activities in the field of multirotor UAVs. More the ones offered at Politecnico di Milano in view of direct relevance to the course topic in technical the direct relevance to the course topic in technical at providing teams of students the opportunity to group carry out design activities in the field of multirotor UAVs. More the direct relevance to the course topic in technical precisely, the course, aimed at an interdisciplinary directoffered relevance to the course topic inintechnical precisely, the course, aimed at an interdisciplinary group the ones at Politecnico di Milano view of terms. terms. out designthe activities the field of multirotor UAVs. group More precisely, course,inaimed at an interdisciplinary terms. of students students comprising Master students in Aeronautical Aeronautical of comprising Master students in the direct relevance to the course topic in technical terms. ••• Introductory Introductory lectures were prepared to provide all of students Master students in Aeronautical precisely, thecomprising course, aimed at an interdisciplinary group Introductory lectures lectures were were prepared prepared to to provide provide all all Engineering, Space Engineering, Automation and Control Engineering, Space Engineering, Automation and Control Control Engineering, Space Engineering, Automation and • terms. Introductory lectures with were aaprepared to provide on all participating students common background participating students with common background on of students comprising Master students in Aeronautical Engineering, Space Engineering, Automation and Control participating students with a common background on Engineering and and Computer Computer Engineering, Engineering, covers covers the the whole whole participating students with a common background on Engineering • Introductory lectures were prepared to each provide all multirotor so to that multirotor UAVs, so as to ensure that each of the Engineering, Space Engineering, Automation and Engineering and Engineering, covers theControl whole multirotor UAVs, UAVs, so as as to aensure ensure that each of of the the design cycle cycle for for a Computer multirotor UAV, from conceptual conceptual design multirotor UAVs, so as to ensure that each of the design a multirotor UAV, from design participating students with common background on teams would be able to work together on UAV design teams would be able to work together on UAV design design cycle for a multirotor UAV, from conceptual design Engineering and Computer Engineering, covers the whole teams wouldUAVs, be able toaswork together on UAV design to in-flight in-flight validation, validation, with specific specific reference to modelling, modelling, to with reference to multirotor soto to ensure that each ofwere the teams would be able work together on UAV design problems. Exercises in multirotor UAV sizing problems. Exercises in multirotor UAV sizing were to in-flight with specific reference to modelling, design cyclevalidation, for a multirotor UAV, from conceptual design problems. Exercises in multirotor UAV sizing were simulation, identification and control. As will be discussed simulation, identification and control. As will be discussed teams would be able to work together on UAV design simulation, identification and control. As will be discussed problems. Exercises in multirotor UAV sizing were also carried out to make sure the students actually also out sure actually to in-flight validation, with specific reference simulation, identification and control. Asbeen willto bemodelling, discussed also carried carriedExercises out to to make make sure the the students students actually in the the following sections, the course has been conceived as also carried out to make sure the students actually in following sections, the course has conceived as problems. in multirotor UAV (see sizing were grasped the overall design methodology Section grasped the overall design methodology (see Section simulation, identification andcourse control. Asbeen will be discussed in the following sections, the has conceived as grasped the overall design methodology (see Section an extra-curricular activity, taking place outside regular grasped the overall designsure methodology (see actually Section an extra-curricular activity, taking place outside regular also carried out to make the students 4 for details). 4 for details). an extra-curricular activity, taking place outside regular in thehours following the coursesohas conceived as 4 for details). class and sections, during weekends, weekends, thebeen emphasis is not class hours so the emphasis is the overall design methodology to (seebeSection 4Subsequently, for details). ••• grasped three Subsequently, three design specifications impleclass hours and and during during weekends, soplace thehands-on emphasis is not not an extra-curricular activity, taking outside regular Subsequently, three design design specifications specifications to to be be impleimpleon conventional lectures but rather on experion conventional lectures but rather on hands-on experi4Subsequently, for details). on conventional lectures but rather on hands-on experi• three design specifications to be implemented by the students were presented in detail and mented by the students were presented in detail and class hours and during weekends, so the emphasis is not on conventional lectures but rather on hands-on experimented by the students were presented in detail and ence in in hardware/software hardware/software integration, integration, data data collection collection and and mented by the students were presented in detail and ence • Subsequently, three design specifications to be implestudents were allocated to corresponding interdiscistudents were allocated to corresponding interdiscion conventional lectures but rather ondata hands-on experience in hardware/software integration, collection and students were allocated to corresponding interdiscianalysis and flight testing. analysis and mented bywere the teams. students were presented in detail and students allocated to corresponding interdisciplinary plinary design teams. analysis and flight flight testing. testing. integration, data collection and ence in hardware/software plinary design design teams. students were allocated to corresponding interdisciplinary design teams. • Each of the teams carried out a preliminary design, The paper paper isflight organized as follows. follows. The The approach approach to to the the •• Each the carried analysis andis testing. The organized as Each of of design the teams teams carried out out aa preliminary preliminary design, design, teams. • plinary Each of the teams carried outpresented a preliminary design, The paper isof as is follows. The in approach the using the methods and tools in initial organisation oforganized the course course is presented in Section to 2 and and using the methods and tools presented in the initial organisation the presented Section 2 using the methods and tools presented in the thedesign, initial organisation the course presented Section to 2 and using of thethe methods and tools in the initial • Each teams carried outpresented a preliminary The paper isoforganized as is follows. The in approach the the methods and tools presented in the initial organisation of the course presented in Section 2 and Copyright © 2019 2019 IFAC 490 2405-8963 Copyright © 2019. TheisAuthors. Published by Elsevier Ltd. 490 All rights using reserved. Copyright © IFAC Copyright © 2019 IFAC 490 Peer review©under of International Federation of Automatic Copyright 2019 responsibility IFAC 490 Control. 10.1016/j.ifacol.2019.11.291 Copyright © 2019 IFAC 490
2019 IFAC ACA August 27-30, 2019. Cranfield, UK
•
•
•
•
• •
Mattia Giurato et al. / IFAC PapersOnLine 52-12 (2019) 490–495
part of the course. Such designs were subsequently reviewed by the instructors. Detailed designs were then carried out, mainly by the student teams, with some feedback from the instructors. The outputs of the detailed designs, corresponding to CAD drawings of the multirotors and corresponding bills of materials, were then subjected to reviews. Having reached an appropriate maturity for the designs, the components (flight control and companion computers, blades, motors, electronic speed controllers, batteries, materials for mechanical integration) needed for platform integration were acquired (or, in some cases such as carbon-fiber frames, manufactured to design). Experimental characterisation of the platforms: in the design of multirotor UAVs most of the modelling uncertainty is associated with the propulsion subsystem (Electronic Speed Controllers (ESCs), motors and propellers) so dedicated data-collection experiments were carried out to characterise such subsystems (see Section 6) for details. Customisation of multirotor simulation: the numerical values of the parameters obtained from the previous activity were used to customise a general-purpose MATLAB/Simulink simulation model for multirotor UAVs to suit each of the three designed platforms. Integration: the student teams then took care (with some support from the instructors) of the mechanical, electrical and electronic integration of the platforms. Flight-testing: flight-testing was used to fine-tune the controller parameters starting from the ones determined in simulation. Finally, endurance tests were carried out to compare actual to required performance.
The course activities have taken place in the Politecnico di Milano Flying Arena for Rotorcraft Technologies (FlyART, see Figure 1), a facility which has been designed to support research and education activities in the field of multirotor UAVs, both for single platforms and formations, with specific reference to guidance, navigation and control systems. More precisely, Fly-ART includes an indoor flight-test facility with a 290m3 flight space covered by a 3D motion capture system, a few work stations for hardware integration and a classroom which can seat up to 25 students.
491
common ground on multirotor UAVs, their principles of operation, architecture and main characteristics from the point of view of preliminary design. In detail, the following topics were covered (see also Table 1): • Course introduction and overview of multirotor UAVs: the first lectures were aimed at providing some basic information about the organisation of the course and, mostly, a general overview of multirotor UAVs, in terms of basic principles of operation, dynamic modelling (rigid body, motors, propellers), simulation, model identification and control. Note that all the involved students have a sound background in dynamic systems, linear control theory and parameter estimation, so that the above topics could be covered in a very efficient way. • Subsystem decomposition and modelling for sizing: two lectures were devoted to the illustration of the main subsystems into which a typical multirotor UAV can be decomposed, namely frame, propulsive system, power supply, electronics and payload. For each subsystem the key parameters playing a role in the sizing were highlighted. • Formulation of sizing problems and development of a simple sizing tool: a simple approach to the sizing of a multirotor UAV (see Section 4 for details) was then presented and the students were asked to both implement their own version of the sizing algorithm and test it using predefined numerical examples. • Introduction to eCalc: for validation purposes, the online multirotor sizing tool eCalc (see Solution for All Markus Mueller (2019)) was also presented and used to double-check the results of the numerical examples. Topic Course introduction Overview of multirotor UAVs Subsystem decomposition Modelling for sizing Formulation of sizing problem Development of sizing tool Introduction to ecalc Presentation of design specifications
Lectures (h) 1 1 1 1 1 1 1 1
Table 1. Syllabus for lectures.
4. QUADROTOR DESIGN 4.1 Design approach
Fig. 1. The FlyART facility at Politecnico di Milano. 3. SYLLABUS FOR INTRODUCTORY LECTURES The introductory lectures mentioned in the previous section were aimed at providing the students an appropriate 491
The design approach used in the framework of the UAV Lab course is very simple and is based on the following considerations and assumptions. The flight time is computed considering a hovering static flight condition. Clearly, in a real flight scenario the flight time will be smaller, according to flight speed, environmental conditions etc. Aerodynamic considerations are neglected at this preliminary level. Furthermore note that, if needed, a size constraint requirement could be considered during the components selection phase. Also, since a specific thrust value can be produced by many motor/propeller pairs, the right choice is considered as the one closer to the required use: in general, a bigger rotor is also more efficient. The
2019 IFAC ACA 492 August 27-30, 2019. Cranfield, UK
Mattia Giurato et al. / IFAC PapersOnLine 52-12 (2019) 490–495
procedure can be summarised as follows (see also Figure 2): • define high level requirements for Maximum Take-Off Weight (MTOW) and endurance (i.e., flight time). • Translate the high level requirements to physical quantities, i.e., maximum thrust and energy. • Select the components of the UAV in order to satisfy the given requirements in terms of thrust and energy. • Verify by analysis that the solution is feasible and close to the initial requirements.
Parameter TOW Flight time Landing pad size Payload
Value As light as possible > 15min > xxm2 > 1kg
Table 2. Design requirements for the carrier multirotor. Parameter TOW Flight time Frame size Payload
Value 0.6kg < TOW < 0.8kg > 10min 250mm FPV camera and video antenna
Table 3. Design requirements for the lander multirotor.
Fig. 2. Block diagram of the design approach. 4.2 Design requirements The students were divided in three teams, making sure that each group had the required mutidisciplinary character aimed for from the outset of the initiative. Three sets of design requirements were then provided to the students. The specified designs were defined based on recent and ongoing research activities within the Aerospace Systems and Control Laboratory (ASCL) of Politecnico di Milano, specifically on the problem of air-to-air landing of multirotor UAVs (see Giuri et al. (2019)) and on the design, prototyping and control of thrust-vectoring multirotor UAVs (see Invernizzi and Lovera (2018)). The problem of air-to-air refuelling is well-known and can arise when undertaking long-range flights. In the military field, Air-to-Air Automatic Refuelling (AAAR) involving fixed-wing drones is object of studies and research activities. Also small UAVs suffer from low endurance problems, since the overwhelming majority of them has an electric propulsion system. A possibility to extend the range of UAV missions could be to have a carrier drone, possibly a fixed-wing one, with several lightweight multirotors aboard, which can take-off from and land on it. The study of automatic air-to-air landing requires the availability of two custom-designed platforms: • a carrier drone, specifically designed to be as insensitive as possible to the perturbations caused by landing and to offer a wide, flat, ”landing-pad-like” surface to carry out landing experiments in a simple and safe way; • a lightweight and agile drone, to be used as a lander. For the lander a requirement specification inspired by high-agility First-Person View (FPV) racing drones has been proposed to the students. In view of this, the design requirements summarised in Table 2 and in Table 3 were formulated. As for thrust-vectoring multirotor UAVs: in recent years the development of multirotor UAVs with thrust vectoring capabilities has received a growing interest. These systems 492
can achieve a larger degree of actuation compared to coplanar multirotor UAVs since both thrust and torque can be oriented within the airframe. This feature makes thrustvectoring UAVs capable of performing complex full-pose maneuvers, which is particularly attractive for inspectionlike applications that may require, for instance, navigation in a constrained environment. Moreover, being able to deliver both force and torque in any direction enhances the UAV interaction capabilities with the environment, which is especially desirable in aerial manipulation tasks. Two main technological solutions have been proposed to endow multirotor UAVs with thrust vectoring capabilities: by employing tiltable propellers Ryll et al. (2015); Kastelan et al. (2015); Invernizzi et al. (2018) and by mounting the propellers in a fixed, non-coplanar fashion Crowther et al. (2011); Rajappa et al. (2015); Brescianini and D’Andrea (2016). In the UAV Lab course one of the student teams was asked to propose a design for a thrustvectoring multirotor UAV belonging to the first class. The main points of the corresponding design specification, summarised in Table 4 therefore require that the UAV includes independent tilting mechanisms for each of the arms, to be treated as a payload in the mass budget of the UAV. Parameter TOW Flight time Thrust-to-weight ratio Payload
Value 0.4kg < TOW < 1kg > 10min >3 Tilting mechanism treated as payload
Table 4. Design requirements for the tiltrotor UAV. 5. STUDENT DESIGN ACTIVITIES Starting from the lectures described in Section 3 and the design approach and requirements outlined in Section 4, the students worked on the second part of the course, the intended planning of which is summarised in Table 5. As can be seen from the table, the course planning required the students to first use the requirements as a main driver to the definition of the configuration and the sizing of the platform, in terms of endurance, takeoff weight etc.. Having established the main configuration parameters, the students then moved to the detailed design, focusing on the mechanical and electrical aspects, placing of the components and wiring. Subsequently, following acquisition of the components for the construction
2019 IFAC ACA August 27-30, 2019. Cranfield, UK
Mattia Giurato et al. / IFAC PapersOnLine 52-12 (2019) 490–495
Activity Platform sizing Platform design Component testing and characterisation Platform simulation model and tool Control oriented models Platform integration and characterisation Control law tuning Test-bed verification In-flight verification Preparation of final report Preparation of final presentation
493
propulsive subsystems, their characterisation was carried out using the setup illustrated in Figure 5, which was already available within the laboratory. The setup allows to measure the thrust and the angular rate produced by the propeller, so as to collect data such as the ones depicted in Figure 6 and Figure 7 for, respectively, the ω vs throttle and thrust vs ω characteristics, where ω is the propeller angular rate. The same figures also report linear and quadratic models obtained from the two datasets. Such models, together with inertial properties derived from CAD drawings, were then used to tune a generalpurpose MATLAB/Simulink quadrotor simulation model which was provided to the students.
Duration (h) 2 3 4 2 2 4 4 2 2 4 4
Table 5. Student activities with planned durations. of the multirotors, the students carried out the mechanical, electrical and electronic integration tasks and proceeded to the characterisation of the propulsion subsystems and the calibration of the simulation model (see the following section for further details). In the actual implementation of the planning in Table 5, however, integration activities turned out to be significantly more time-consuming than anticipated, so that test-bed verification was skipped and controller setup had to be reduced to simple in-flight tuning based on empirical rules, prior to the execution of the final endurance tests to validate the designs against the initial requirements. The final tasks carried out by the student teams consisted in the preparation of a design report and of a presentation of the results, followed by a technical discussion.
Fig. 4. CAD drawing of the lander multirotor.
6. OBTAINED RESULTS In this section the results obtained by the UAV Lab students are presented and discussed, with specific reference to one of the three teams, for the sake of conciseness. More precisely, the focus will be on the design of the lander drone, starting from the requirements listed in Table 3. Using the sizing approach outlined in Section 4, the lander team was able to work out a sizing and a set of components compatible with the required MTOW and endurance. For validation purposes, the obtained solution was analysed using eCalc; the results of the eCalc analysis in terms of range and endurance, reported in Figure 3, were fully consistent with the preliminary sizing.
Fig. 5. Setup for drive system characterisation.
Fig. 3. Performance characteristics of the final design for the lander. The CAD drawing of the designed lander multirotor is depicted in Figure 4. Following the integration of the 493
Finally, following the complete integration of the platform (see Figure 8) it was possible to verify compliance with the original requirements. Besides TOW, which is easily checked, to verify the expected performance in terms of hover endurance dedicated tests were carried out. The results in terms of TOW and endurance are reported in Table 6, from which it can be seen that the designed multirotor is compliant with the original requirements. Finally, time histories of the measured attitude and position (under feedback control) during the endurance tests are depicted in Figure 9 and Figure 10. As can be seen, even though limited time was available for the tuning of the attitude and position control loops, accurate pointing and positioning were achieved.
2019 IFAC ACA 494 August 27-30, 2019. Cranfield, UK
Mattia Giurato et al. / IFAC PapersOnLine 52-12 (2019) 490–495
Fig. 9. Hover control performance: attitude errors.
Fig. 6. Drive system characterisation: ω vs throttle.
Fig. 10. Hover control performance: position control. Fig. 7. Drive system characterisation: thrust vs ω.
7. CONCLUSIONS In this paper, an outline of the UAV Lab multirotor design and integration course has been presented and discussed. As described in the previous sections, the course emphasized hands-on experience with respect to conventional lectures, leveraging the available competences of the students and the multidisciplinary nature of the teams. The experience of the first iteration on this course has been extremely positive from the students’ point of view, both in terms of direct feedback to the instructors and in terms of evaluations gathered anonymously through suitable forms. In particular, the design and built multirotors are now being used for research activities within FLyART. In this respect the UAV Lab course turned out to be an effective form of synergy between education and research. For future iterations of the course the schedule will be revised to take into account the past experience, namely the very time-consuming nature of integration activities, the original planning of which did not consider to a sufficient extent the students’ learning curve in this task. Indeed this went to the detriment of the foreseen activities in control design, which had to be reduced to fit the overall schedule.
Fig. 8. The integrated lander UAV. Parameter TOW Flight time Frame size
Requirement 0.6kg < TOW < 0.8kg > 10min 250mm
Design 0.683kg 10.5min 250mm
Result 0.734kg 13.3min 250mm
8. ACKNOWLEDGEMENTS
Table 6. Design requirements, design results and experimental results.
The Authors would like to thank the students who participated in the course for their enthusiasm and passion. 494
2019 IFAC ACA August 27-30, 2019. Cranfield, UK
Mattia Giurato et al. / IFAC PapersOnLine 52-12 (2019) 490–495
This paper is dedicated to the memory of one our students, Pietro Rapini, who passed away suddenly in December 2018. REFERENCES Brescianini, D. and D’Andrea, R. (2016). Design, modeling and control of an omni-directional aerial vehicle. In Proc. IEEE Int. Conf. Robotics and Automation (ICRA), 3261–3266. doi:10.1109/ICRA.2016.7487497. Crowther, B., Lanzon, A., Maya-Gonzalez, M., and Langkamp, D. (2011). Kinematic analysis and control design for a nonplanar multirotor vehicle. Journal of Guidance, Control, and Dynamics, 34(4), 1157–1171. Gaponov, I. and Razinkova, A. (2012). Quadcopter design and implementation as a multidisciplinary engineering course. In IEEE International Conference on Teaching, Assessment and Learning for Engineering (TALE). Giuri, P., Marini Cossetti, A., Giurato, M., and Lovera, M. (2019). Air-to-air automatic landing for multirotor UAVs. In 5th CEAS Conference on Guidance, Navigation and Control, Milano, Italy. Invernizzi, D., Giurato, M., Gattazzo, P., and Lovera, M. (2018). Full pose tracking for a tilt-arm quadrotor UAV. In Proc. IEEE Conf. Control Technology and Applications (CCTA), 159–164. doi: 10.1109/CCTA.2018.8511566. Invernizzi, D. and Lovera, M. (2018). Trajectory tracking control of thrust vectoring UAVs. Automatica, 95, 180– 186. Kastelan, D., Konz, M., and Rudolph, J. (2015). Fully actuated tricopter with pilot-supporting control. IFACPapersOnLine, 48(9), 79–84. Khan, S., Jaffery, M.H., Hanif, A., and Asif, M.R. (2017). Teaching tool for a control systems laboratory using a quadrotor as plant in MATLAB. IEEE Transactions on Education, 60(4), 249–256. Rajappa, S., Ryll, M., B¨ ulthoff, H.H., and Franchi, A. (2015). Modeling, control and design optimization for a fully-actuated hexarotor aerial vehicle with tilted propellers. In 2015 IEEE International Conference on Robotics and Automation, 4006–4013. IEEE. Ryll, M., B¨ ulthoff, H.H., and Robuffo Giordano, P. (2015). A novel overactuated quadrotor unmanned aerial vehicle: Modeling, control, and experimental validation. IEEE Transactions on Control Systems Technology, 23(2), 540–556. Solution for All Markus Mueller (2019). eCalc. https://www.ecalc.ch/.
495
495
ScienceDirect
IFAC PapersOnLine 52-12 (2019) 490–495
UAV UAV UAV UAV
Lab: Lab: Lab: Lab:
a a a a
multidisciplinary UAV design multidisciplinary multidisciplinary UAV UAV design design course multidisciplinary UAV design course course ∗ ∗ Mattia Giurato ∗ Paolo Gattazzo ∗∗∗ Marco Lovera ∗∗∗
Mattia Mattia Giurato Giurato ∗ Paolo Paolo Gattazzo Gattazzo ∗ Marco Marco Lovera Lovera ∗ Mattia Giurato ∗ Paolo Gattazzo ∗ Marco Lovera ∗ ∗ ∗ ∗ ∗ Mattia Giurato Paolo Gattazzo Marco Lovera ∗ Dipartimento ∗ di Scienze Tecnologie Aerospaziali, Politecnico di Dipartimento di di Scienze Scienze eee Tecnologie Tecnologie Aerospaziali, Aerospaziali, Politecnico Politecnico di di ∗ Dipartimento ∗ Dipartimento di Scienze e 34, Tecnologie Aerospaziali, Politecnico di Milano, Via La Masa 20156, Milano, Italy (e-mail: Milano, Via La Masa 34, 20156, Milano, Italy (e-mail: Milano, Via La Masa 34, 20156, Milano, Italy (e-mail: ∗ Milano, Via La Masa 34, 20156, Milano, Italy (e-mail: Dipartimento di Scienze e Tecnologie Aerospaziali, Politecnico di [email protected]). [email protected]). [email protected]). Milano, Via [email protected]). Masa 34, 20156, Milano, Italy (e-mail: [email protected]). Abstract: Aerospace Aerospace control control education education can can significantly significantly benefit benefit from from actual actual hands-on hands-on experiexperiAbstract: Abstract: Aerospace control education can significantly benefit fromto actual hands-on experience. In most cases, however, such experience can only be provided to students in small-scale ence. In most cases, however, such experience can only be provided students in small-scale ence. In most cases, however, such experience can only bebenefit provided students in small-scale Abstract: Aerospace control can significantly fromto actual hands-on experience. Inactivities. most cases, however, such experience provided to students in small-scale project activities. In this this papereducation the experience experience ofcan theonly UAVbe Lab course, held at Politecnico Politecnico for the the project In paper the of the UAV Lab course, held at for ence. Inactivities. most cases, however, such experience can only be provided to students in small-scale project In this paper the experience of the UAV Lab course, held at Politecnico for the first time in the fall of 2018, is presented and discussed. The course, aimed at an interdisciplinary first time in of 2018, is presented and discussed. The course, aimed at an interdisciplinary first time in the the fall fallIn ofthis 2018, presented and discussed. TheLab course, aimed an interdisciplinary project paper the experience offor theaa UAV course, heldatat Politecnico for the group ofactivities. students, covers theiswhole whole design cycle for multirotor UAV, from conceptual design to group of students, covers the whole design cycle multirotor UAV, from conceptual design to group of students, covers the design cycle for a multirotor UAV, from conceptual design to first time in the fall of 2018, is presented and discussed. The course, aimed at an interdisciplinary group of students, covers the whole design cycle for a multirotor UAV, from conceptual design to in-flight validation, with specific reference to modelling, simulation, identification and control. in-flight validation, with specific reference to modelling, simulation, identification and control. in-flight with the specific reference to modelling, simulation, identification anddesign control. group of validation, students, covers whole cycle for a multirotor UAV, from to in-flight validation, with specific reference to modelling, simulation, The course course has been been conceived asdesign an extra-curricular extra-curricular activity, soidentification the conceptual emphasisand is control. not on on The has conceived as an activity, so the emphasis is not in-flight validation, with specific reference to experience modelling, in simulation, identification and control. The course has been conceived as an extra-curricular activity, so the emphasis is not on conventional lectures but rather on hands-on hardware/software integration, data conventional lectures but rather on experience in hardware/software integration, data conventional lectures rather onashands-on hands-on experience inactivity, hardware/software integration, The course beenbut conceived an The extra-curricular so syllabus the emphasis is notdata on collection andhas analysis and flight testing. testing. The paper presents the course course syllabus and organisation organisation collection and analysis and flight testing. paper presents the course and organisation collection and analysis and flight The paper presents the syllabus and conventional lectures but rather on hands-on experience in hardware/software integration, data collection and analysis and flight testing. The paper presents the course syllabus and organisation and provides provides an an overview overview of of the the obtained obtained results results and and the the feedback feedback provided provided by by the the students. students. and provides an overview of the obtained results and the feedback provided by the students. and collection andan analysis andof flight paperand presents the course syllabus and provides overview the testing. obtainedThe results the feedback provided byand theorganisation students. Copyright © 2019. Authors.of Published by Elsevier Ltd. and All rights reserved. provided by the students. and provides anThe overview the obtained results the feedback 1. INTRODUCTION the 1. INTRODUCTION INTRODUCTION the corresponding syllabus is discussed in Section 3. Sub1. the corresponding corresponding syllabus syllabus is is discussed discussed in in Section Section 3. 3. SubSub1. INTRODUCTION the corresponding syllabus is discussed in Section 3.teams Subsequently, the sizing approach used by the student sequently, the sizing approach used by the student teams sequently, the sizing approach used by the student teams 1. INTRODUCTION the corresponding syllabus is discussed in Section 3.teams Subsequently, the approach usedactivities by the student to carry carry out thesizing conceptual design activities and the the three out the conceptual design and three In recent recent years years Unmanned Unmanned Aerial Aerial Vehicles Vehicles (UAVs), (UAVs), which which to sequently, the sizing approach usedactivities by the student student teams to carry out the conceptual design and the three In requirement specifications provided to the teams specifications provided to the student teams In recent years Aerialonly Vehicles (UAVs), which requirement in the past had been for applicarequirement specifications provided to an theoverview student to carry out the conceptual design activities and theteams three in the the past had Unmanned been of interest only for military military applicaare then illustrated in Section 4, while of in past had been of of interest interest only for military applicaare then illustrated in Section 4, while an overview of the are then illustrated in Section 4, while an overview of the the In recent years Unmanned Aerial Vehicles (UAVs), which in the past had been of interest only for military applications, have have started started to to play play aa significant significant role role in in civil civil applica- requirement specifications provided to the student teams are then illustrated in Section 4, while an overview of the tions, students’ design activities and an example of the obtained students’ design activities and an example of the obtained students’ design activities and an example of the obtained in theas past had been of interest onlyand for military tions, have started to play a significant role in civil applications well, ranging from personal commercial use to are then(referring illustrated in Section 4,one while ofmulthe students’ design activities andto example ofdesigned the obtained tions as well, ranging from and commercial use to results specifically toan of an theoverview designed specifically one of the multions as well, ranging fromapersonal personal commercial to results tions, have started toapplications. play significant in civil applicacountless industrial In and therole framework ofuse civil results (referring (referring specifically of the mulstudents’ design activities andtoanone example ofdesigned the 5obtained countless industrial applications. In the framework of civil tirotors) are provided, respectively, in Section and in are provided, respectively, in Section 55 and in countless industrial In and the framework ofuse civil tions as well, rangingapplications. from personal commercial to tirotors) applications, multirotor UAVs represent the most common tirotors)(referring are provided, respectively, inaathe Section and in results specifically to one of designed mulapplications, multirotor UAVs represent the most most common common Section 6. Finally Section 7 provides discussion of the applications, multirotor UAVs represent the Section 6. Finally Section 7 provides discussion of the Finally Section 7 providesina Section discussion of the countless industrial In the framework of As civil Section 6.are applications, multirotor UAVs represent the most common architecture, due to toapplications. their versatility and reliability. provided, respectively, 5 and in Section 6. Finally Section 7 provides a discussion of the architecture, due their versatility and reliability. As aa tirotors) overall course course experience, experience, some some lessons lessons learned learned and and aa few few applications, UAVs represent most common architecture, due to their versatility andthe reliability. As of consequence, multirotor education activities related to the design design ofa overall Section 6. Finally Section 7 provides discussion the overall course experience, some lessonsalearned and of a few consequence, education activities related to the perspectives for further developments. perspectives for further developments. consequence, education activities related to the design of architecture, due have to their versatility andmore reliability. As a overall multirotor UAVs UAVs become more and and widespread, perspectives further developments. coursefor experience, some lessons learned and a few multirotor have become more more widespread, multirotor UAVs have both become more and more widespread, consequence, education activities related to the design of of perspectives for further developments. with courses covering specific disciplinary aspects with courses courses covering both specific disciplinary aspects of with covering both specific disciplinary aspects of 2. APPROACH APPROACH 2. multirotor UAVs have become more and more widespread, with courses covering both specific disciplinary aspects of their design design and and operation operation (aeromechanics, (aeromechanics, power power electronelectrontheir 2. APPROACH with courses covering both(aeromechanics, specific disciplinary of their design and operation poweraspects electronics, hardware hardware and software, navigation, control, teleme2. APPROACH ics, and software, navigation, control, telemeThe The course was organised and managed according to the ics, hardware software, navigation, control, telemetheir design andand operation (aeromechanics, power electronThe course course was was organised organised and and managed managed according according to to the the try/communications etc.) and system-level design issues try/communications etc.) and system-level design issues The course was organised and managed according to the following approach: following approach: ics, hardware and software, navigation, control, telemetry/communications etc.) and system-level design issues following approach: (see, e.g., Gaponov and Razinkova (2012) Khan et al. (see, e.g., Gaponov and Razinkova (2012) Khan et al. course was organised and managed according to the (see, e.g., Gaponov and Razinkova (2012) Khan issues et al. The following approach: try/communications and system-level (see, e.g., Gaponov etc.) and Razinkova (2012) design Khan et al. following (2017)). ••• aaa call for call for the definition of three interdisciplinary (2017)). approach: call for the the definition definition of of three three interdisciplinary interdisciplinary (see, e.g., Gaponov and Razinkova (2012) Khan et al. (2017)). • ateams call was for the definition ofstudents three interdisciplinary sent to Master in teams was sent to Master students in Aeronautical teams was sent to Master students in Aeronautical Aeronautical In this paper the experience of the UAV Lab course, In this paper the experience of the UAV Lab course, (2017)). • ateams call for the definition of three interdisciplinary In this paper the experience of the UAV Lab course, was sent to Master students in Aeronautical Engineering, Space Engineering, Computer Science Engineering, Space Engineering, Computer Science In this paper the experience of the UAV Lab course, Engineering, Space Engineering, Computer Science held at Politecnico di Milano for the first time in the Engineering, Space Engineering, Computer Science held at Politecnico di Milano for the first time in the teams was sent to Master students in Aeronautical and Engineering, Automation and Control Engineerand Engineering, Automation and Control EngineerIn this paper the experience of the UAV Lab course, held at Politecnico di Milano for the first time in the and Engineering, Automation and Control Engineerfall of 2018, is presented and discussed. The course aims fall of is presented and The course aims Engineering, Space Engineering, Computer Science and Engineering, Automation and Control Engineering. four programs were selected among ing. These four Master programs were selected among fall ofat2018, 2018, isteams presented and discussed. discussed. The time course aims held Politecnico di students Milano for first the ing. These These four Master Master programs were selected among at providing of the opportunity to carry at providing providing teams of students the the opportunity toincarry carry and Engineering, Automation and Control Engineerat teams of students the opportunity to ing. These four Master programs were selected among the ones offered at Politecnico di Milano in view of the ones offered at Politecnico di Milano in view of fall of 2018, is presented and discussed. The course aims at providing teams of students the opportunity to carry the ones offered at Politecnico di Milano in view of out design activities in the field of multirotor UAVs. More out design activities in the field of multirotor UAVs. More ing. These four Master programs were selected among out design activities in the field of multirotor UAVs. More the ones offered at Politecnico di Milano in view of direct relevance to the course topic in technical the direct relevance to the course topic in technical at providing teams of students the opportunity to group carry out design activities in the field of multirotor UAVs. More the direct relevance to the course topic in technical precisely, the course, aimed at an interdisciplinary directoffered relevance to the course topic inintechnical precisely, the course, aimed at an interdisciplinary group the ones at Politecnico di Milano view of terms. terms. out designthe activities the field of multirotor UAVs. group More precisely, course,inaimed at an interdisciplinary terms. of students students comprising Master students in Aeronautical Aeronautical of comprising Master students in the direct relevance to the course topic in technical terms. ••• Introductory Introductory lectures were prepared to provide all of students Master students in Aeronautical precisely, thecomprising course, aimed at an interdisciplinary group Introductory lectures lectures were were prepared prepared to to provide provide all all Engineering, Space Engineering, Automation and Control Engineering, Space Engineering, Automation and Control Control Engineering, Space Engineering, Automation and • terms. Introductory lectures with were aaprepared to provide on all participating students common background participating students with common background on of students comprising Master students in Aeronautical Engineering, Space Engineering, Automation and Control participating students with a common background on Engineering and and Computer Computer Engineering, Engineering, covers covers the the whole whole participating students with a common background on Engineering • Introductory lectures were prepared to each provide all multirotor so to that multirotor UAVs, so as to ensure that each of the Engineering, Space Engineering, Automation and Engineering and Engineering, covers theControl whole multirotor UAVs, UAVs, so as as to aensure ensure that each of of the the design cycle cycle for for a Computer multirotor UAV, from conceptual conceptual design multirotor UAVs, so as to ensure that each of the design a multirotor UAV, from design participating students with common background on teams would be able to work together on UAV design teams would be able to work together on UAV design design cycle for a multirotor UAV, from conceptual design Engineering and Computer Engineering, covers the whole teams wouldUAVs, be able toaswork together on UAV design to in-flight in-flight validation, validation, with specific specific reference to modelling, modelling, to with reference to multirotor soto to ensure that each ofwere the teams would be able work together on UAV design problems. Exercises in multirotor UAV sizing problems. Exercises in multirotor UAV sizing were to in-flight with specific reference to modelling, design cyclevalidation, for a multirotor UAV, from conceptual design problems. Exercises in multirotor UAV sizing were simulation, identification and control. As will be discussed simulation, identification and control. As will be discussed teams would be able to work together on UAV design simulation, identification and control. As will be discussed problems. Exercises in multirotor UAV sizing were also carried out to make sure the students actually also out sure actually to in-flight validation, with specific reference simulation, identification and control. Asbeen willto bemodelling, discussed also carried carriedExercises out to to make make sure the the students students actually in the the following sections, the course has been conceived as also carried out to make sure the students actually in following sections, the course has conceived as problems. in multirotor UAV (see sizing were grasped the overall design methodology Section grasped the overall design methodology (see Section simulation, identification andcourse control. Asbeen will be discussed in the following sections, the has conceived as grasped the overall design methodology (see Section an extra-curricular activity, taking place outside regular grasped the overall designsure methodology (see actually Section an extra-curricular activity, taking place outside regular also carried out to make the students 4 for details). 4 for details). an extra-curricular activity, taking place outside regular in thehours following the coursesohas conceived as 4 for details). class and sections, during weekends, weekends, thebeen emphasis is not class hours so the emphasis is the overall design methodology to (seebeSection 4Subsequently, for details). ••• grasped three Subsequently, three design specifications impleclass hours and and during during weekends, soplace thehands-on emphasis is not not an extra-curricular activity, taking outside regular Subsequently, three design design specifications specifications to to be be impleimpleon conventional lectures but rather on experion conventional lectures but rather on hands-on experi4Subsequently, for details). on conventional lectures but rather on hands-on experi• three design specifications to be implemented by the students were presented in detail and mented by the students were presented in detail and class hours and during weekends, so the emphasis is not on conventional lectures but rather on hands-on experimented by the students were presented in detail and ence in in hardware/software hardware/software integration, integration, data data collection collection and and mented by the students were presented in detail and ence • Subsequently, three design specifications to be implestudents were allocated to corresponding interdiscistudents were allocated to corresponding interdiscion conventional lectures but rather ondata hands-on experience in hardware/software integration, collection and students were allocated to corresponding interdiscianalysis and flight testing. analysis and mented bywere the teams. students were presented in detail and students allocated to corresponding interdisciplinary plinary design teams. analysis and flight flight testing. testing. integration, data collection and ence in hardware/software plinary design design teams. students were allocated to corresponding interdisciplinary design teams. • Each of the teams carried out a preliminary design, The paper paper isflight organized as follows. follows. The The approach approach to to the the •• Each the carried analysis andis testing. The organized as Each of of design the teams teams carried out out aa preliminary preliminary design, design, teams. • plinary Each of the teams carried outpresented a preliminary design, The paper isof as is follows. The in approach the using the methods and tools in initial organisation oforganized the course course is presented in Section to 2 and and using the methods and tools presented in the initial organisation the presented Section 2 using the methods and tools presented in the thedesign, initial organisation the course presented Section to 2 and using of thethe methods and tools in the initial • Each teams carried outpresented a preliminary The paper isoforganized as is follows. The in approach the the methods and tools presented in the initial organisation of the course presented in Section 2 and Copyright © 2019 2019 IFAC 490 2405-8963 Copyright © 2019. TheisAuthors. Published by Elsevier Ltd. 490 All rights using reserved. Copyright © IFAC Copyright © 2019 IFAC 490 Peer review©under of International Federation of Automatic Copyright 2019 responsibility IFAC 490 Control. 10.1016/j.ifacol.2019.11.291 Copyright © 2019 IFAC 490
2019 IFAC ACA August 27-30, 2019. Cranfield, UK
•
•
•
•
• •
Mattia Giurato et al. / IFAC PapersOnLine 52-12 (2019) 490–495
part of the course. Such designs were subsequently reviewed by the instructors. Detailed designs were then carried out, mainly by the student teams, with some feedback from the instructors. The outputs of the detailed designs, corresponding to CAD drawings of the multirotors and corresponding bills of materials, were then subjected to reviews. Having reached an appropriate maturity for the designs, the components (flight control and companion computers, blades, motors, electronic speed controllers, batteries, materials for mechanical integration) needed for platform integration were acquired (or, in some cases such as carbon-fiber frames, manufactured to design). Experimental characterisation of the platforms: in the design of multirotor UAVs most of the modelling uncertainty is associated with the propulsion subsystem (Electronic Speed Controllers (ESCs), motors and propellers) so dedicated data-collection experiments were carried out to characterise such subsystems (see Section 6) for details. Customisation of multirotor simulation: the numerical values of the parameters obtained from the previous activity were used to customise a general-purpose MATLAB/Simulink simulation model for multirotor UAVs to suit each of the three designed platforms. Integration: the student teams then took care (with some support from the instructors) of the mechanical, electrical and electronic integration of the platforms. Flight-testing: flight-testing was used to fine-tune the controller parameters starting from the ones determined in simulation. Finally, endurance tests were carried out to compare actual to required performance.
The course activities have taken place in the Politecnico di Milano Flying Arena for Rotorcraft Technologies (FlyART, see Figure 1), a facility which has been designed to support research and education activities in the field of multirotor UAVs, both for single platforms and formations, with specific reference to guidance, navigation and control systems. More precisely, Fly-ART includes an indoor flight-test facility with a 290m3 flight space covered by a 3D motion capture system, a few work stations for hardware integration and a classroom which can seat up to 25 students.
491
common ground on multirotor UAVs, their principles of operation, architecture and main characteristics from the point of view of preliminary design. In detail, the following topics were covered (see also Table 1): • Course introduction and overview of multirotor UAVs: the first lectures were aimed at providing some basic information about the organisation of the course and, mostly, a general overview of multirotor UAVs, in terms of basic principles of operation, dynamic modelling (rigid body, motors, propellers), simulation, model identification and control. Note that all the involved students have a sound background in dynamic systems, linear control theory and parameter estimation, so that the above topics could be covered in a very efficient way. • Subsystem decomposition and modelling for sizing: two lectures were devoted to the illustration of the main subsystems into which a typical multirotor UAV can be decomposed, namely frame, propulsive system, power supply, electronics and payload. For each subsystem the key parameters playing a role in the sizing were highlighted. • Formulation of sizing problems and development of a simple sizing tool: a simple approach to the sizing of a multirotor UAV (see Section 4 for details) was then presented and the students were asked to both implement their own version of the sizing algorithm and test it using predefined numerical examples. • Introduction to eCalc: for validation purposes, the online multirotor sizing tool eCalc (see Solution for All Markus Mueller (2019)) was also presented and used to double-check the results of the numerical examples. Topic Course introduction Overview of multirotor UAVs Subsystem decomposition Modelling for sizing Formulation of sizing problem Development of sizing tool Introduction to ecalc Presentation of design specifications
Lectures (h) 1 1 1 1 1 1 1 1
Table 1. Syllabus for lectures.
4. QUADROTOR DESIGN 4.1 Design approach
Fig. 1. The FlyART facility at Politecnico di Milano. 3. SYLLABUS FOR INTRODUCTORY LECTURES The introductory lectures mentioned in the previous section were aimed at providing the students an appropriate 491
The design approach used in the framework of the UAV Lab course is very simple and is based on the following considerations and assumptions. The flight time is computed considering a hovering static flight condition. Clearly, in a real flight scenario the flight time will be smaller, according to flight speed, environmental conditions etc. Aerodynamic considerations are neglected at this preliminary level. Furthermore note that, if needed, a size constraint requirement could be considered during the components selection phase. Also, since a specific thrust value can be produced by many motor/propeller pairs, the right choice is considered as the one closer to the required use: in general, a bigger rotor is also more efficient. The
2019 IFAC ACA 492 August 27-30, 2019. Cranfield, UK
Mattia Giurato et al. / IFAC PapersOnLine 52-12 (2019) 490–495
procedure can be summarised as follows (see also Figure 2): • define high level requirements for Maximum Take-Off Weight (MTOW) and endurance (i.e., flight time). • Translate the high level requirements to physical quantities, i.e., maximum thrust and energy. • Select the components of the UAV in order to satisfy the given requirements in terms of thrust and energy. • Verify by analysis that the solution is feasible and close to the initial requirements.
Parameter TOW Flight time Landing pad size Payload
Value As light as possible > 15min > xxm2 > 1kg
Table 2. Design requirements for the carrier multirotor. Parameter TOW Flight time Frame size Payload
Value 0.6kg < TOW < 0.8kg > 10min 250mm FPV camera and video antenna
Table 3. Design requirements for the lander multirotor.
Fig. 2. Block diagram of the design approach. 4.2 Design requirements The students were divided in three teams, making sure that each group had the required mutidisciplinary character aimed for from the outset of the initiative. Three sets of design requirements were then provided to the students. The specified designs were defined based on recent and ongoing research activities within the Aerospace Systems and Control Laboratory (ASCL) of Politecnico di Milano, specifically on the problem of air-to-air landing of multirotor UAVs (see Giuri et al. (2019)) and on the design, prototyping and control of thrust-vectoring multirotor UAVs (see Invernizzi and Lovera (2018)). The problem of air-to-air refuelling is well-known and can arise when undertaking long-range flights. In the military field, Air-to-Air Automatic Refuelling (AAAR) involving fixed-wing drones is object of studies and research activities. Also small UAVs suffer from low endurance problems, since the overwhelming majority of them has an electric propulsion system. A possibility to extend the range of UAV missions could be to have a carrier drone, possibly a fixed-wing one, with several lightweight multirotors aboard, which can take-off from and land on it. The study of automatic air-to-air landing requires the availability of two custom-designed platforms: • a carrier drone, specifically designed to be as insensitive as possible to the perturbations caused by landing and to offer a wide, flat, ”landing-pad-like” surface to carry out landing experiments in a simple and safe way; • a lightweight and agile drone, to be used as a lander. For the lander a requirement specification inspired by high-agility First-Person View (FPV) racing drones has been proposed to the students. In view of this, the design requirements summarised in Table 2 and in Table 3 were formulated. As for thrust-vectoring multirotor UAVs: in recent years the development of multirotor UAVs with thrust vectoring capabilities has received a growing interest. These systems 492
can achieve a larger degree of actuation compared to coplanar multirotor UAVs since both thrust and torque can be oriented within the airframe. This feature makes thrustvectoring UAVs capable of performing complex full-pose maneuvers, which is particularly attractive for inspectionlike applications that may require, for instance, navigation in a constrained environment. Moreover, being able to deliver both force and torque in any direction enhances the UAV interaction capabilities with the environment, which is especially desirable in aerial manipulation tasks. Two main technological solutions have been proposed to endow multirotor UAVs with thrust vectoring capabilities: by employing tiltable propellers Ryll et al. (2015); Kastelan et al. (2015); Invernizzi et al. (2018) and by mounting the propellers in a fixed, non-coplanar fashion Crowther et al. (2011); Rajappa et al. (2015); Brescianini and D’Andrea (2016). In the UAV Lab course one of the student teams was asked to propose a design for a thrustvectoring multirotor UAV belonging to the first class. The main points of the corresponding design specification, summarised in Table 4 therefore require that the UAV includes independent tilting mechanisms for each of the arms, to be treated as a payload in the mass budget of the UAV. Parameter TOW Flight time Thrust-to-weight ratio Payload
Value 0.4kg < TOW < 1kg > 10min >3 Tilting mechanism treated as payload
Table 4. Design requirements for the tiltrotor UAV. 5. STUDENT DESIGN ACTIVITIES Starting from the lectures described in Section 3 and the design approach and requirements outlined in Section 4, the students worked on the second part of the course, the intended planning of which is summarised in Table 5. As can be seen from the table, the course planning required the students to first use the requirements as a main driver to the definition of the configuration and the sizing of the platform, in terms of endurance, takeoff weight etc.. Having established the main configuration parameters, the students then moved to the detailed design, focusing on the mechanical and electrical aspects, placing of the components and wiring. Subsequently, following acquisition of the components for the construction
2019 IFAC ACA August 27-30, 2019. Cranfield, UK
Mattia Giurato et al. / IFAC PapersOnLine 52-12 (2019) 490–495
Activity Platform sizing Platform design Component testing and characterisation Platform simulation model and tool Control oriented models Platform integration and characterisation Control law tuning Test-bed verification In-flight verification Preparation of final report Preparation of final presentation
493
propulsive subsystems, their characterisation was carried out using the setup illustrated in Figure 5, which was already available within the laboratory. The setup allows to measure the thrust and the angular rate produced by the propeller, so as to collect data such as the ones depicted in Figure 6 and Figure 7 for, respectively, the ω vs throttle and thrust vs ω characteristics, where ω is the propeller angular rate. The same figures also report linear and quadratic models obtained from the two datasets. Such models, together with inertial properties derived from CAD drawings, were then used to tune a generalpurpose MATLAB/Simulink quadrotor simulation model which was provided to the students.
Duration (h) 2 3 4 2 2 4 4 2 2 4 4
Table 5. Student activities with planned durations. of the multirotors, the students carried out the mechanical, electrical and electronic integration tasks and proceeded to the characterisation of the propulsion subsystems and the calibration of the simulation model (see the following section for further details). In the actual implementation of the planning in Table 5, however, integration activities turned out to be significantly more time-consuming than anticipated, so that test-bed verification was skipped and controller setup had to be reduced to simple in-flight tuning based on empirical rules, prior to the execution of the final endurance tests to validate the designs against the initial requirements. The final tasks carried out by the student teams consisted in the preparation of a design report and of a presentation of the results, followed by a technical discussion.
Fig. 4. CAD drawing of the lander multirotor.
6. OBTAINED RESULTS In this section the results obtained by the UAV Lab students are presented and discussed, with specific reference to one of the three teams, for the sake of conciseness. More precisely, the focus will be on the design of the lander drone, starting from the requirements listed in Table 3. Using the sizing approach outlined in Section 4, the lander team was able to work out a sizing and a set of components compatible with the required MTOW and endurance. For validation purposes, the obtained solution was analysed using eCalc; the results of the eCalc analysis in terms of range and endurance, reported in Figure 3, were fully consistent with the preliminary sizing.
Fig. 5. Setup for drive system characterisation.
Fig. 3. Performance characteristics of the final design for the lander. The CAD drawing of the designed lander multirotor is depicted in Figure 4. Following the integration of the 493
Finally, following the complete integration of the platform (see Figure 8) it was possible to verify compliance with the original requirements. Besides TOW, which is easily checked, to verify the expected performance in terms of hover endurance dedicated tests were carried out. The results in terms of TOW and endurance are reported in Table 6, from which it can be seen that the designed multirotor is compliant with the original requirements. Finally, time histories of the measured attitude and position (under feedback control) during the endurance tests are depicted in Figure 9 and Figure 10. As can be seen, even though limited time was available for the tuning of the attitude and position control loops, accurate pointing and positioning were achieved.
2019 IFAC ACA 494 August 27-30, 2019. Cranfield, UK
Mattia Giurato et al. / IFAC PapersOnLine 52-12 (2019) 490–495
Fig. 9. Hover control performance: attitude errors.
Fig. 6. Drive system characterisation: ω vs throttle.
Fig. 10. Hover control performance: position control. Fig. 7. Drive system characterisation: thrust vs ω.
7. CONCLUSIONS In this paper, an outline of the UAV Lab multirotor design and integration course has been presented and discussed. As described in the previous sections, the course emphasized hands-on experience with respect to conventional lectures, leveraging the available competences of the students and the multidisciplinary nature of the teams. The experience of the first iteration on this course has been extremely positive from the students’ point of view, both in terms of direct feedback to the instructors and in terms of evaluations gathered anonymously through suitable forms. In particular, the design and built multirotors are now being used for research activities within FLyART. In this respect the UAV Lab course turned out to be an effective form of synergy between education and research. For future iterations of the course the schedule will be revised to take into account the past experience, namely the very time-consuming nature of integration activities, the original planning of which did not consider to a sufficient extent the students’ learning curve in this task. Indeed this went to the detriment of the foreseen activities in control design, which had to be reduced to fit the overall schedule.
Fig. 8. The integrated lander UAV. Parameter TOW Flight time Frame size
Requirement 0.6kg < TOW < 0.8kg > 10min 250mm
Design 0.683kg 10.5min 250mm
Result 0.734kg 13.3min 250mm
8. ACKNOWLEDGEMENTS
Table 6. Design requirements, design results and experimental results.
The Authors would like to thank the students who participated in the course for their enthusiasm and passion. 494
2019 IFAC ACA August 27-30, 2019. Cranfield, UK
Mattia Giurato et al. / IFAC PapersOnLine 52-12 (2019) 490–495
This paper is dedicated to the memory of one our students, Pietro Rapini, who passed away suddenly in December 2018. REFERENCES Brescianini, D. and D’Andrea, R. (2016). Design, modeling and control of an omni-directional aerial vehicle. In Proc. IEEE Int. Conf. Robotics and Automation (ICRA), 3261–3266. doi:10.1109/ICRA.2016.7487497. Crowther, B., Lanzon, A., Maya-Gonzalez, M., and Langkamp, D. (2011). Kinematic analysis and control design for a nonplanar multirotor vehicle. Journal of Guidance, Control, and Dynamics, 34(4), 1157–1171. Gaponov, I. and Razinkova, A. (2012). Quadcopter design and implementation as a multidisciplinary engineering course. In IEEE International Conference on Teaching, Assessment and Learning for Engineering (TALE). Giuri, P., Marini Cossetti, A., Giurato, M., and Lovera, M. (2019). Air-to-air automatic landing for multirotor UAVs. In 5th CEAS Conference on Guidance, Navigation and Control, Milano, Italy. Invernizzi, D., Giurato, M., Gattazzo, P., and Lovera, M. (2018). Full pose tracking for a tilt-arm quadrotor UAV. In Proc. IEEE Conf. Control Technology and Applications (CCTA), 159–164. doi: 10.1109/CCTA.2018.8511566. Invernizzi, D. and Lovera, M. (2018). Trajectory tracking control of thrust vectoring UAVs. Automatica, 95, 180– 186. Kastelan, D., Konz, M., and Rudolph, J. (2015). Fully actuated tricopter with pilot-supporting control. IFACPapersOnLine, 48(9), 79–84. Khan, S., Jaffery, M.H., Hanif, A., and Asif, M.R. (2017). Teaching tool for a control systems laboratory using a quadrotor as plant in MATLAB. IEEE Transactions on Education, 60(4), 249–256. Rajappa, S., Ryll, M., B¨ ulthoff, H.H., and Franchi, A. (2015). Modeling, control and design optimization for a fully-actuated hexarotor aerial vehicle with tilted propellers. In 2015 IEEE International Conference on Robotics and Automation, 4006–4013. IEEE. Ryll, M., B¨ ulthoff, H.H., and Robuffo Giordano, P. (2015). A novel overactuated quadrotor unmanned aerial vehicle: Modeling, control, and experimental validation. IEEE Transactions on Control Systems Technology, 23(2), 540–556. Solution for All Markus Mueller (2019). eCalc. https://www.ecalc.ch/.
495
495