- Email: [email protected]
200 Manual control of slowly responding processes by use of online simulation models
Abstracts
749
197 DAISY, A Knowledgeable Monitoring and Warning Aid for the Driver on German Motorways M. Kopf, R. Onken, pp 195-200
201 Aiding the Operator in the Manual Control of a Space Manipulator J.F.T. Bos, H.G. Stassen, A. van Lunteren, pp 215-220
Design and implementation of the driver monitoring and warning system DAISY (Driver Assisting SYstem) are presented. A mathematical formulation of danger is based upon a parameter space description of traffic situations in terms of the time reserve with respect to danger-avoiding action. There are three main modules: the reference driver module, the model of the actual driver, and the discrepancy interpretation module. In a learning phase DAISY exlzacts the actual driver's driving characteristics in order to adapt the warnings later on. The system is currently implemented in a well-equipped fixed-base driving simulator. Some experimental results are reported. Parts of the system will be ported to an experimental car.
An operator manually controlling a space manipulator is faced with several difficulties. Three problems were investigated: lack of direct vision, up to 6 degrees-offreedom (IX)F) to be controlled, and slow and complex dynamics. The results showed that reference lines enhance the 3D perception; further automation of the insert task led to a more-efficient task execution; and the display of the stopping configuration can lead to a faster and safer task execution. The display of the predicted trajectory led to less energy consumption due to a calmer control behaviour. The length of the prediction horizon should equal half of the stopping time.
202 Biomechanics in Aircraft Control M.M. van Paassen, pp 221-226 198 Knowledge Based Cockpit Assistant for Controlled Airspace Flight Operation T. Wittig, R. Onken, pp 195-200 A knowledge-based cockpit assistant for flight operation under instrument Flight Rules (IFR) is presented. The system is aimed at improvement of situation assessment, including monitoring of the pilot crew activities and performance increase by computer aids for flight planning and plan execution. The modular system structure is described, as well as the individual system modules. A first prototype was tested about two years ago in a flight simulator at the University of the German Armed Forces in Munich by professional pilots under realistic IFR-scenarios. The results of these test runs, proving enhancements in overall system performance, are presented.
This paper presents a linear model for a pilot's neuromuscular system. This is then combined with models of the pilot's crossover behaviour, the side stick and aircraft roll control. With this model set, three control situations with side sticks are analyzed. The model is compared to data from an experiment with the side stick. The factors contributing to roll ratchet are studied by a root-locus analysis. Control with an active side stick is considered. The model structure corresponds to that of the neuromuscular system, so the parameters known for parts of the neuromuscular system, such as muscle data. may be used in the model.
203 Analysis and Modelling of Pilot Airplane Interaction by an Integrated Simulation Approach P.C. Cacclabue, G. Cojazzi, S. Mancini,
199 A Connectionist Traffic Sign Recognition System for Onboard Driver Information D. Krumblegel, K.-F. Kraiss, S. Schreiber, pp 201-206 This paper deals with the development of a support system for vehicle handling based on a traffic sign reco.gnition system. This pattern recognition system consists of three modules dedicated to detection, recognition, and knowledge-based control. Neural networks are used as adaptive classifiers. The stage of development and the applicability of such a system are discussed.
E. Hollnagei, pp 227-234 This paper presents a preliminary simplified application of the System Response Generator (SRG) concept, focusing on the dynamic aspects of pilot-ahplane interaction and on pilot errors. Three components play a role: the deterministic simulation of the airplane and its control mechanisms; the simulation of the p,ilot behaviour; and a model for failure/error generation, l w o simulations are described in detail. The error module has been based on a simple generator of systematic events of omissions of proceduralised actions; hardware failures are not considered. The results are compared with the data obtained performing a 'classical' human reliability study for the same case, with very positive f'mdings.
200 Manual Control of Slowly Responding Processes by Use of Online Simulation Models C. Hessler, pp 207-214
204 Method for the Probabilistic Failure Analysis of a Man-Machine System G. Heslinga, pp 235-242
Online simulation models can be used as operator aids for the control of slowly responding processes. The operator uses a quick motion simulation, run in parallel to the ~eal process, to evaluate the effect of future control inputs. A new graphical interactive man-machine interface allows easy manipulation of the model. In the control input mode the operator evaluates the effects of different control sequences, and in the trajectory input mode inversion of the model is used for the computation of appropriate conu'ol sequences. These have been applied suecessfuUy to a ship-manoeuvring task and to a load-distribution task in a combined heat and power station in laboratory studies.
Probabilistic failure analyses of complex installations are usually made for a plant in a steady state. The current techniques are however too limited for non-steady state situations (e.g. in starting up a plant). This paper presents a possible technique for evaluating the safety of a man-machine system that is brought to another state for which procedural actions have to be performed. It combines a method for the identification of the systems that may occur due to human errors in following the procedure incorrectly, and a systems analysis of these system states. An example related to a man-machine system for locking out electrical power for maintenance is presented.
749
197 DAISY, A Knowledgeable Monitoring and Warning Aid for the Driver on German Motorways M. Kopf, R. Onken, pp 195-200
201 Aiding the Operator in the Manual Control of a Space Manipulator J.F.T. Bos, H.G. Stassen, A. van Lunteren, pp 215-220
Design and implementation of the driver monitoring and warning system DAISY (Driver Assisting SYstem) are presented. A mathematical formulation of danger is based upon a parameter space description of traffic situations in terms of the time reserve with respect to danger-avoiding action. There are three main modules: the reference driver module, the model of the actual driver, and the discrepancy interpretation module. In a learning phase DAISY exlzacts the actual driver's driving characteristics in order to adapt the warnings later on. The system is currently implemented in a well-equipped fixed-base driving simulator. Some experimental results are reported. Parts of the system will be ported to an experimental car.
An operator manually controlling a space manipulator is faced with several difficulties. Three problems were investigated: lack of direct vision, up to 6 degrees-offreedom (IX)F) to be controlled, and slow and complex dynamics. The results showed that reference lines enhance the 3D perception; further automation of the insert task led to a more-efficient task execution; and the display of the stopping configuration can lead to a faster and safer task execution. The display of the predicted trajectory led to less energy consumption due to a calmer control behaviour. The length of the prediction horizon should equal half of the stopping time.
202 Biomechanics in Aircraft Control M.M. van Paassen, pp 221-226 198 Knowledge Based Cockpit Assistant for Controlled Airspace Flight Operation T. Wittig, R. Onken, pp 195-200 A knowledge-based cockpit assistant for flight operation under instrument Flight Rules (IFR) is presented. The system is aimed at improvement of situation assessment, including monitoring of the pilot crew activities and performance increase by computer aids for flight planning and plan execution. The modular system structure is described, as well as the individual system modules. A first prototype was tested about two years ago in a flight simulator at the University of the German Armed Forces in Munich by professional pilots under realistic IFR-scenarios. The results of these test runs, proving enhancements in overall system performance, are presented.
This paper presents a linear model for a pilot's neuromuscular system. This is then combined with models of the pilot's crossover behaviour, the side stick and aircraft roll control. With this model set, three control situations with side sticks are analyzed. The model is compared to data from an experiment with the side stick. The factors contributing to roll ratchet are studied by a root-locus analysis. Control with an active side stick is considered. The model structure corresponds to that of the neuromuscular system, so the parameters known for parts of the neuromuscular system, such as muscle data. may be used in the model.
203 Analysis and Modelling of Pilot Airplane Interaction by an Integrated Simulation Approach P.C. Cacclabue, G. Cojazzi, S. Mancini,
199 A Connectionist Traffic Sign Recognition System for Onboard Driver Information D. Krumblegel, K.-F. Kraiss, S. Schreiber, pp 201-206 This paper deals with the development of a support system for vehicle handling based on a traffic sign reco.gnition system. This pattern recognition system consists of three modules dedicated to detection, recognition, and knowledge-based control. Neural networks are used as adaptive classifiers. The stage of development and the applicability of such a system are discussed.
E. Hollnagei, pp 227-234 This paper presents a preliminary simplified application of the System Response Generator (SRG) concept, focusing on the dynamic aspects of pilot-ahplane interaction and on pilot errors. Three components play a role: the deterministic simulation of the airplane and its control mechanisms; the simulation of the p,ilot behaviour; and a model for failure/error generation, l w o simulations are described in detail. The error module has been based on a simple generator of systematic events of omissions of proceduralised actions; hardware failures are not considered. The results are compared with the data obtained performing a 'classical' human reliability study for the same case, with very positive f'mdings.
200 Manual Control of Slowly Responding Processes by Use of Online Simulation Models C. Hessler, pp 207-214
204 Method for the Probabilistic Failure Analysis of a Man-Machine System G. Heslinga, pp 235-242
Online simulation models can be used as operator aids for the control of slowly responding processes. The operator uses a quick motion simulation, run in parallel to the ~eal process, to evaluate the effect of future control inputs. A new graphical interactive man-machine interface allows easy manipulation of the model. In the control input mode the operator evaluates the effects of different control sequences, and in the trajectory input mode inversion of the model is used for the computation of appropriate conu'ol sequences. These have been applied suecessfuUy to a ship-manoeuvring task and to a load-distribution task in a combined heat and power station in laboratory studies.
Probabilistic failure analyses of complex installations are usually made for a plant in a steady state. The current techniques are however too limited for non-steady state situations (e.g. in starting up a plant). This paper presents a possible technique for evaluating the safety of a man-machine system that is brought to another state for which procedural actions have to be performed. It combines a method for the identification of the systems that may occur due to human errors in following the procedure incorrectly, and a systems analysis of these system states. An example related to a man-machine system for locking out electrical power for maintenance is presented.