- Email: [email protected]
Improvement of boom control performance for hybrid hydraulic excavator with potential energy recovery
Automation in Construction 30 (2013) 161–169
Contents lists available at SciVerse ScienceDirect
Automation in Construction journal homepage: www.elsevier.com/locate/autcon
Improvement of boom control performance for hybrid hydraulic excavator with potential energy recovery Tao Wang ⁎, Qingfeng Wang, Tianliang Lin The State Key Laboratory of Fluid Power Transmission and Control, Zhejiang University, 310027 Hangzhou, China
a r t i c l e
i n f o
Article history: Accepted 7 November 2012 Available online 13 December 2012 Keywords: Energy recovery Excavator Hybrid system Boom operability Hydraulic motor Permanent magnet generator Control strategy
a b s t r a c t Potential energy recovery (ER) is an effective way to reduce energy consumption of hybrid hydraulic excavators; however, the ER system with a direct speed-control strategy is prone to oscillation of actuators due to the reduction of damping in comparison to the conventional throttle governing system. This paper aims to improve the boom control performance of a hybrid hydraulic excavator by properly designing the ER controller. Based on the dynamics of the system, mathematical modeling including hydraulic components and electrical components is carried out. A staged composite control strategy is proposed to achieve acceptable performance in the whole velocity range. Load torque observation is employed to increase the speed stiffness of the permanent magnet generator applied in the system as well as the stiffness of the boom motion. The leakage flow which affects the anti-disturbance capability and accuracy of the control system is compensated. Finally, the effectiveness of the proposed control scheme is verified by simulation and experimental results. Crown Copyright © 2012 Published by Elsevier B.V. All rights reserved.
1. Introduction Various technologies are developed to reduce energy consumption and emission with the raising demand of energy saving and environmental protection in the world range. Hybrid power system, which has been successfully applied to vehicles, is also paid much attention in the field of construction machinery, especially hydraulic excavators [1–6]. The hybrid system usually has two energy sources including a combustion engine and an electrical energy storage device. The fuel economy is improved by operating the engine in an optimum efficiency range with a proper control strategy. According to the structures of hybrid systems, hybrid hydraulic excavators (HHEs) can be categorized as series type, parallel type and compound type, of which the compound type hybrid structure is the best solution from the aspects of the fuel efficiency, the additional cost and the expected payback time [6]. In hydraulic systems, energy recovery (ER) is another energy saving method which can be realized by using hydraulic or electrical energy storage devices. The hydraulic approach is to convert the recoverable energy to hydraulic form, store it in a hydraulic accumulator and release it when there are requirements [7–9]. However, it needs additional components such as hydraulic pump/motors or transformers to reuse the recovered energy. The electrical approach, which converts the recoverable energy to electrical form, is more suitable for HHEs, for the hybrid power systems provide electrical energy storage devices such as batteries or super capacitors and the recovered energy can be delivered directly to
⁎ Corresponding author. Tel.: +86 571 87951314 8108; fax: +86 571 87951941. E-mail address: [email protected] (T. Wang).
any electrical actuators. In general, there are two kinds of ER systems in HHEs, including the braking kinetic energy of swings and the gravitational potential energy of booms. The former is similar to the energy regeneration system in hybrid or electrical vehicles which have been studied widely, so this paper concentrates on the latter. A potential energy recovery system for hydraulic forklift trucks was developed and verified in [10]. In comparison to the conventional machines, the efficiency was increased from 56% to 74% at high velocity and from 39% to 69% at low velocity, but the boom response was more oscillatory. In [11], a similar system was studied and the efficiency evaluation of every component is carried out by theoretical analysis and experimental tests. It is reported that the maximum recovery efficiency was 66.2% and improvements could still be achieved. Researches have also been carried out on the HHEs. In [12], additional hydraulic motor-generator was installed in the return oil line of the boom circuit to recover the potential energy, for most hydraulic pumps could not be used in motoring mode. The results of experiment and simulation showed that the ER scheme was feasible in HHEs. Two different ER configurations were compared in [13]. The simulation results showed that the configuration with a hydraulic accumulator had better control performance; however, it was much more complex and its efficiency was only 41% due to the extra link of energy conversion in the accumulator. This paper studies the boom control performance of the HHE with a potential ER system where an electrical generator is driven by a hydraulic motor in the return oil line. The control characteristics of the novel system are analyzed based on the dynamic model. In order to improve the comprehensive performance, solutions including composite control strategy, load torque observation, and leakage flow compensation are
0926-5805/$ – see front matter. Crown Copyright © 2012 Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.autcon.2012.11.034
162
T. Wang et al. / Automation in Construction 30 (2013) 161–169
proposed. The effectiveness of the proposed control scheme is verified by simulations and experiments. The paper is organized as follows. Section 2 presents the structure and parameters of the potential ER system and the experimental platform. Mathematical modeling and analysis are demonstrated in Section 3. Section 4 concentrates on the controller design with the assist of simulations in the MATLAB environment. Section 5 provides the experimental results. Finally, conclusions are drawn in Section 6.
2. Structure and parameters 2.1. System structure The configuration of the compound type HHE with a potential ER system is shown in Fig. 1. It can be seen that the hydraulic pump is driven by the engine and the electrical machine together. The electrical machine behaves as a motor when the load is heavy and a generator when the load is light. By this way, the working condition of the engine is improved and the fuel consumption is reduced. The conventional hydraulic swing motor is replaced with the electrical motor which can be used to recover the braking kinetic energy. The capacitor's state of
charge (SOC) is restrained by a dynamic-working-point control developed in [2]. The boom potential ER system is composed of a hydraulic motor, an electrical generator and a throttle valve. The gravitational potential energy can be converted to the electrical form and stored in the super capacitor when the boom is being lowered down, instead of being dissipated totally in the conventional throttle governing mode. In the proposed mode, the boom velocity is adjusted by controlling the rotational speed of the generator as well as the hydraulic motor. It should be noticed that the throttle valve is necessary in the system, for the oil line should be cut off when the boom is lifted up. Furthermore, the valve can be used to improve the dynamic performance at starting and the low velocity stage, as described in Section 4. Generally, the energy conversion efficiency and the actuator control performance are two important evaluation indexes of the ER system. The total efficiency is mainly determined by the efficiencies of the recovering components and the hydraulic circuits. The control performance indicates the actuator responses to the movement commands which are usually conveyed through joysticks by drivers. Due to the introduction of the ER system, the operation mode becomes different from the conventional mode. Therefore, the velocity governing performance is also changed and should be reconsidered.
Fig. 1. Configuration of the compound HHEs with potential ER.
T. Wang et al. / Automation in Construction 30 (2013) 161–169
2.2. Experimental platform
163
Table 1 Parameters of the experimental platform.
As shown in Fig. 2, an experimental platform is established to study the potential ER system for HHEs. The boom mechanism of a seven-ton hydraulic excavator is employed. The type selection of the key components of the ER system should follow the principles of high efficiency, good control performance and high power density. Thus we apply an axial piston hydraulic motor and a permanent magnet synchronous generator to the platform. The generator is connected with an insulated gate bipolar transistor (IGBT) rectifier. An encoder is used to detect the rotor position and speed information for close-loop control. A DS1104 digital signal processing card is used as the controller where control algorithms are carried out. There are some additional instruments such as boom displacement sensor, torque sensor, pressure sensor and flow meter which are employed to detect the corresponding variables in real time, and these variables are not used for feedback control. The main parameters of the experimental platform are presented in Table 1.
Component
Parameter
Value
Boom cylinder
Rod diameter (mm) Piston diameter (mm) Stroke length (m) Rated flow (L/min) Displacement (mL) Maximum speed (rpm) Rated torque (Nm) Maximum torque (Nm) Maximum current (A) Maximum voltage (V) Capacity (F) Rated voltage (V)
65 115 0.75 100 55 2100 100 160 200 600 6.25 400
Proportional directional valve Hydraulic motor Electrical generator
Rectifier Super capacitor
chambers of the cylinder, A0 and A1 are the corresponding working areas, f is the coulomb friction and Bc is the coefficient of viscous friction. The flow equation of the rodless chamber of the cylinder is derived as
3. Mathematical modeling The first step is to investigate the control characteristics of the ER process with mathematical tools. When the boom is being lowered down, the dynamic equation of the hydraulic cylinder can be given as mc
dvc ¼ F g þ P 0 A0 −P 1 A1 −f −Bc vc dt
ð1Þ
where mc is the equivalent mass of load, vc is the cylinder velocity, Fg is the load force induced by gravity, P0 and P1 are the pressures in the two
V 1 dP 1 ¼ A1 vc −Q v −C 1 P 1 βe dt
ð2Þ
where V1 is the volume of the chamber, βe is the effective bulk modulus of hydraulic oil, C1 is leakage coefficient and Qv is the valve flow which can be expressed as sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2ðP 1 −P 2 Þ Q v ¼ C d Wxv ρ
Fig. 2. Schematic of the experimental platform.
ð3Þ
164
T. Wang et al. / Automation in Construction 30 (2013) 161–169
coefficient. The dynamics of the motor and the generator can be given as Jt
dωm P 2 Dm ¼ η −T e −Bm ωm dt 2π m
ð8Þ
where Jt is the total moment of inertia, ηm is the mechanical efficiency, Te is the electromagnetic torque of the generator and Bm is the coefficient of viscous friction. The permanent magnet generator is a non-salient pole machine, so Te can be expressed as [14] Te ¼
3 pψ i 2 f q
ð9Þ
where p is the number of pole pairs, ψf is the permanent magnet flux linkage and iq is the q-axis current. The voltage equations of the generator in the dq-coordinate are written as Fig. 3. Leakage coefficient of the hydraulic motor.
where Cd is the flow coefficient, W is the area gradient of the valve orifice, xv is the spool displacement, P2 is the pressure in the output port of the valve and ρ is the density of hydraulic oil. The dynamic equation of the spool is obtained by
F m ¼ mv
d2 xv dx þ Bv v þ K v xv dt dt 2
ð4Þ
where Fm is the output force of the proportional solenoid, mv is the mass of the spool, Bv is the coefficient of viscous friction and Kv is the stiffness of spring. The value of Fm is almost linear to the coil current iv and can be expressed as F m ¼ K f iv
ð5Þ
where Kf is the force-current gain. The current dynamics in the solenoid coil is derived as uv ¼ Lv
div dx þ Rv iv þ K b v dt dt
ð6Þ
where uv is the output voltage of the amplifier and is proportional to the control signal, Lv is the inductance and Rv is the resistance of the coil, Kb is the coefficient of the back electromotive force. The flow equation of the inlet chamber of the hydraulic motor is obtained by V 2 dP 2 ω D ¼ Q v − m m −C 2 P 2 βe dt 2π
ð7Þ
where V2 is the volume of the chamber, Dm is the displacement, ωm is the angular speed of the hydraulic motor, and C2 is leakage
ud ¼ −Rs id −Ls
did þ pωm Ls iq dt
ð10Þ
uq ¼ −Rs iq −Ls
diq −pωm Ls id þ pωm ψf dt
ð11Þ
where ud and uq are the output voltages of the vector-controlled rectifier, id is the d-axis current, Rs is the resistance and Ls is the inductance of the winding. According to the equations above, it can be seen that the cylinder velocity is controlled by the opening of the directional valve and the rotational speed of the generator together. A direct strategy is to open the valve fully no matter how much the control command is so as to minimize the losses dissipated in the throttle, and then adjust Te to govern the rotational speed of the generator as well as the cylinder velocity accordingly. However, this strategy is prone to be impactive and oscillatory when the spool moves to the maximum displacement suddenly at starting. In addition, several points should be taken into consideration in this control system. Firstly, the performance of slight operation becomes worse, for the hydraulic motor usually does not work well at the low speed stage. Secondly, the inherent leakage flow of the hydraulic motor affects the anti-disturbance ability and the accuracy of the system. Fig. 3 shows the tested leakage coefficient under various speed of the hydraulic motor. Thirdly, the control link from the generator to the cylinder is an open loop, so the system stiffness is greatly influenced by the close-loop speed stiffness of the generator. Therefore, in order to improve the system performance, the controller should be particularly designed based on the system characteristics. 4. Control system design 4.1. Configuration of control system Fig. 4 shows the block diagram of the control system where uc is the input velocity command, Sc is the SOC of the super capacitor, xv⁎ is the input opening signal of the proportional directional valve, ωm⁎
Fig. 4. Block diagram of the velocity control system.
T. Wang et al. / Automation in Construction 30 (2013) 161–169
165
4.2. Composite control strategy Above all, the system controller should satisfy the conditions that Sc and ωm are no more than their maximum allowable values. Otherwise, the generator should be disabled to avoid damage. Under these conditions, Fig. 5 shows the output signals of the direct strategy and the composite strategy separately. In the direct strategy, the directional valve is fully opened and the generator is enabled to govern the boom cylinder from the starting. In the composite strategy, there are three stages where the valve and generator are controlled with different modes. The principles of the composite strategy are illustrated as follows. Firstly, the recoverable energy is too little to recover and the performance of the hydraulic motor is poor at the low speed stage, so it is better to use the proportional directional valve to control the cylinder and disable the generator at the same time. At this stage, the control signal of the valve is given by
xv ¼ K q uc
ð12Þ
where Kq is the flow gain. Secondly, the directional valve should be opened fully at the high speed stage to minimize the pressure drop on the valve and recover as much energy as possible. Simultaneously the generator is enabled and controlled by the speed signal as
ωm ¼
Fig. 5. Output signals. (a) Direct strategy. (b) Composite strategy.
is the input speed signal and g is the enable signal of the generator. Herein the directional valve and the generator are controlled by a certain strategy where the governing modes vary with different speed stages as described in Section 4.2. The values of Sc and ωm are used as input conditions of the algorithm. Three PI controllers are designed to regulate the currents in Eqs. (6), (10) and (11). The speed of the generator is controlled by a PID controller with load estimation and leakage flow compensation as described in Section 4.3. Since the electrical time constant is generally much smaller than the mechanical time constant, the electrical dynamics have little influence on the dynamic performance of the system and can be regarded as proportional components in the design of the speed controller.
2πvc A1 Dm
ð13Þ
Thirdly, there should be a transition stage between the low and the high speed stage, otherwise impacts and oscillations will be brought in when control signals are sent to the spool at the switch point. The speed control signal of the generator is the same as Eq. (13) at the transition stage. It can be seen that the proportional valve and the generator are governed under different modes according to the values of the input commands in the developed strategy. By using the staged method, the proportional valve is gradually opened and the inlet pressure of the hydraulic motor is gradually increased so that impacts can be eliminated. When slight operation is required, the boom movements are controlled smoothly by the valve and the ER system is disabled. When the boom is being operated fast, the generator is used to govern the cylinder velocity and realize energy recovery. 4.3. Rotational speed governing of the generator The rotational speed control diagram of the generator is shown in Fig. 6. As a simple and reliable method, the PID controller is employed in the main loop. The output of the PID controller is given by t
uðt Þ ¼ K p eðt Þ þ K i ∫0 eðτÞdτ þ K d
Fig. 6. Speed control diagram of the generator.
deðt Þ dt
ð14Þ
166
T. Wang et al. / Automation in Construction 30 (2013) 161–169
where e(t) is the error of angular speed, Kp is the proportional coefficient, Ki is the integral coefficient and Kd is the derivative coefficient. The three coefficients can be obtained with Ziegler–Nichols tuning method. It should be ensured that the generator works in generating mode rather than motoring mode which may result in suction phenomenon. Therefore, the q-axis current output is limited by the saturation of which the upper limit is set to the value corresponding to the maximum electromagnetic torque and the lower limit is set to zero. The system performance is prone to be influenced by the pressure fluctuation in the inlet chamber of the hydraulic motor. In order to reduce the influence and increase the speed stiffness of the generator, a load feedforward loop is designed by using a pressure observer as shown in Fig. 6, instead of using a pressure sensor. According to Eq. (8), the inlet pressure can be estimated by D dωm P^ 2 ¼ m T e þ J t þ Bm ωm GðsÞ 2π dt
where the leakage coefficient can be obtained by the lookup table as shown in Fig. 3. Then the leakage flow is compensated by the following expression
ω ˜ m ¼ ωm −
2πQ l Dm
It can be observed from Eq. (17) that the compensated value will become minus when the speed command is small enough. It means that the generator should be operated as a motor to satisfy the low speed command. This phenomenon is energy-consuming and obviously diseconomical. Fortunately, it can be avoided by using the composite
ð15Þ
where Te can be obtained from the control command, ωm can be obtained by the speed feedback and G(s) is a low-pass filter which is used to depress the noise brought in by the derivation of ωm. Since the load is observed, the leakage flow of the hydraulic motor can be calculated as Q l ¼ C ðωm ÞP^ 2
ð16Þ
Fig. 7. Simulation results under different strategies.
ð17Þ
Fig. 8. Experimental results with potential ER (vc = 0.13 m/s).
T. Wang et al. / Automation in Construction 30 (2013) 161–169
167
of the two different control strategies before experiments. Fig. 7 shows the simulation results. It can be seen that the system with the direct strategy has considerable oscillation at starting. However, by using the proposed composite strategy, the performance can be greatly improved. 5. Experimental study 5.1. Control performance Fig. 8 shows the experimental results with potential ER and Fig. 9 provides the results of throttle governing for comparison. The results indicate that the boom control performance under the proposed
Fig. 9. Experimental results using throttle governing (vc = 0.13 m/s).
strategy described above because the generator is disabled at the low speed stage.
4.4. Simulations Based on the mathematical model, simulations of the control system are carried out in the MATLAB environment to verify the performances
Fig. 10. Experimental results at low velocity (vc = 0.013 m/s).
Fig. 11. Comparison of systems without and with load observation.
168
T. Wang et al. / Automation in Construction 30 (2013) 161–169
strategy is close to the performance of throttle governing although there is a little overshoot. The voltage of the super capacitor which represents the SOC is obviously increased in the ER process. Fig. 10 presents the experimental results at low boom velocity. It can be seen that there are ripples when the boom is controlled by the ER system. So it is proper to control the boom by using a valve when the velocity command is very small as demonstrated in Section 4. Fig. 11 gives the boom velocities of the systems without and with load torque observation. As shown in Fig. 11, the pressures in the rodless chamber of the boom cylinder are varied from 11 MPa to 9 MPa by adjusting the relief valve to simulate the load variation. It can be seen that the velocity fluctuation with load observation is depressed by approximate 50%. The results show that the software method is effective in improving the system stiffness without additional cost of sensors. Fig. 12 gives the boom velocities of the system without and with leakage compensation. It can be found that the accuracy of the compensated system is improved. In fact, the leakage flow of the hydraulic motor is various with the load pressure, so the anti-disturbance capability of the control system can also be enhanced with the compensation. 5.2. Efficiency evaluation A series of experiments with different boom velocities are carried out to investigate the efficiency of the potential ER system. Considering from the aspects of power and energy respectively, the maximum and the average energy conversion efficiencies from the rodless chamber of the boom cylinder to the super capacitor are given as ηec
max
ηec
ave
U c Ic ¼ max Pc Q c ¼
∫U c Ic ∫P c Q c
ηem
max
ave
U c Ic ¼ max Pm Q m ¼
∫U c I c ∫P m Q m
Boom velocity (m/s)
ηem_ave
ηec_ave
ηem_max
ηec_max
0.013 0.025 0.05 0.07 0.10 0.13
0.000 0.549 0.649 0.693 0.662 0.587
0.000 0.467 0.608 0.645 0.542 0.419
0.000 0.756 0.762 0.773 0.804 0.811
0.000 0.635 0.713 0.714 0.657 0.583
where Pm and Qm are the pressure and the flow rate in the inlet chamber of the hydraulic motor. The experimental results are shown in Table 2. The efficiencies are zeros when the boom velocity is very low and the generator is disabled. Relatively high efficiencies can be obtained at the middle velocity stage. It can be deduced that a part of potential energy is dissipated in the directional valve by analyzing the data. Therefore, better results can be achieved by properly selecting the directional valve. The value of ηem_max, which represents the transient energy conversion efficiency of the hydraulic motor and generator, is improved with the increase of the boom velocity, for the generator usually obtains the maximum efficiency when the practical power is equal to its rated power. In general, the potential ER efficiency is acceptable in this experimental platform. 6. Conclusion
ð18Þ
ð19Þ
where Pc and Qc are the pressure and the flow rate in the rodless chamber of the boom cylinder, Uc and Ic is the voltage and the charging current of the super capacitor. The maximum and the average energy conversion efficiencies from the inlet of the hydraulic motor to the super capacitor are given as ηem
Table 2 Test of energy recovery efficiency.
ð20Þ
ð21Þ
The boom control performance of the HHE with a potential ER system is studied in this paper. The characteristics of the control system are analyzed by modeling and simulation. Solutions including composite control strategy, load torque observation and leakage flow compensation are proposed to improve the dynamic performance of the system and maintain relatively high recovery efficiency. The experimental results show that the proposed control scheme can recover the potential energy of the boom effectively with acceptable control performance. The further work is to research the power management when the prototypical potential ER system is applied to a working HHE model. Acknowledgment The authors gratefully acknowledge the financial support of the National Natural Science Foundation of China (Grant No. 50875233) and the National High Technology Research and Development Program of China (Grant No. 2010AA044401). References
Fig. 12. Comparison of systems without and with leakage compensation.
[1] Q.F. Wang, Y.T. Zhang, Q. Xiao, Evaluation for energy saving effect and simulation research on energy saving of hydraulic system in hybrid construction machinery, Chinese Journal of Mechanical Engineering 41 (12) (2005) 135–140. [2] Q. Xiao, Q.F. Wang, Y.T. Zhang, Control strategies of power system in hybrid hydraulic excavator, Automation in Construction 17 (4) (2008) 361–367. [3] M. Ochiai, S. Rye, Hybrid in construction machinery, in: Proceedings of the 7th JFPS International Symposium on Fluid Power, Toyama, 2008, pp. 41–44. [4] D.Y. Wang, C. Guan, S.X. Pan, M.J. Zhang, X. Lin, Performance analysis of hydraulic excavator powertrain hybridization, Automation in Construction 18 (3) (2009) 249–257. [5] J. Choi, H. Kim, S. Yu, K. Yi, Development of integrated controller for a compound hybrid excavator, Journal of Mechanical Science and Technology 25 (6) (2011) 1557–1563. [6] T.-S. Kwon, S.-W. Lee, S.-K. Sul, C.-G. Park, N.-I. Kim, B.-I. Kang, et al., Power control algorithm for hybrid excavator with supercapacitor, IEEE Transactions on Industry Applications 46 (4) (2010) 1447–1455. [7] X.G. Liang, T. Virvalo, Development and research of an energy saving drive in a hydraulic crane, in: The 7th Scandinavian International Conference on Fluid Power, Sweden, 2001, pp. 151–161. [8] T. Virvalo, W. Sun, Improving energy utilization in hydraulic booms - what it is all about, The 6th International Conference on Fluid Power Transmission and Control, Hangzhou, 2005.
T. Wang et al. / Automation in Construction 30 (2013) 161–169 [9] J.D. Van de Ven, M.W. Olson, P.Y. Li, Development of a hydro-mechanical hydraulic hybrid drive train with independent wheel torque control for an urban passenger vehicle, in: Proceedings of the International Fluid Power Exposition, Las Vegas, Nevada, 2008. [10] T.O. Andersen, M.R. Hansen, H.C. Pedersen, F. Conrad, Regeneration of potential energy in hydraulic forklift trucks, in: The 6th International Conference on Fluid Power Transmission and Control, Hangzhou, 2005, pp. 302–306. [11] T. Minav, P. Immonen, L. Laurila, V. Vtorov, J. Pyrhonen, M. Niemela, Electric energy recovery system for a hydraulic forklift – theoretical and experimental evaluation, IET Electric Power Applications 5 (4) (2011) 377–385.
169
[12] Y.T. Zhang, Q.F. Wang, Q. Xiao, Simulation and experimental research on energy regeneration with hydraulic motor for hybrid drive excavator, Chinese Journal of Mechanical Engineering 43 (8) (2007) 218–223. [13] T.L. Lin, Q.F. Wang, B.Z. Hu, W. Gong, Research on the energy regeneration systems for hybrid hydraulic excavators, Automation in Construction 19 (8) (2010) 1016–1026. [14] M. Chinchilla, S. Arnaltes, J.C. Burgos, Control of permanent-magnet generators applied to variable-speed wind-energy systems connected to the grid, IEEE Transactions on Energy Conversion 21 (1) (2006) 130–135.
Contents lists available at SciVerse ScienceDirect
Automation in Construction journal homepage: www.elsevier.com/locate/autcon
Improvement of boom control performance for hybrid hydraulic excavator with potential energy recovery Tao Wang ⁎, Qingfeng Wang, Tianliang Lin The State Key Laboratory of Fluid Power Transmission and Control, Zhejiang University, 310027 Hangzhou, China
a r t i c l e
i n f o
Article history: Accepted 7 November 2012 Available online 13 December 2012 Keywords: Energy recovery Excavator Hybrid system Boom operability Hydraulic motor Permanent magnet generator Control strategy
a b s t r a c t Potential energy recovery (ER) is an effective way to reduce energy consumption of hybrid hydraulic excavators; however, the ER system with a direct speed-control strategy is prone to oscillation of actuators due to the reduction of damping in comparison to the conventional throttle governing system. This paper aims to improve the boom control performance of a hybrid hydraulic excavator by properly designing the ER controller. Based on the dynamics of the system, mathematical modeling including hydraulic components and electrical components is carried out. A staged composite control strategy is proposed to achieve acceptable performance in the whole velocity range. Load torque observation is employed to increase the speed stiffness of the permanent magnet generator applied in the system as well as the stiffness of the boom motion. The leakage flow which affects the anti-disturbance capability and accuracy of the control system is compensated. Finally, the effectiveness of the proposed control scheme is verified by simulation and experimental results. Crown Copyright © 2012 Published by Elsevier B.V. All rights reserved.
1. Introduction Various technologies are developed to reduce energy consumption and emission with the raising demand of energy saving and environmental protection in the world range. Hybrid power system, which has been successfully applied to vehicles, is also paid much attention in the field of construction machinery, especially hydraulic excavators [1–6]. The hybrid system usually has two energy sources including a combustion engine and an electrical energy storage device. The fuel economy is improved by operating the engine in an optimum efficiency range with a proper control strategy. According to the structures of hybrid systems, hybrid hydraulic excavators (HHEs) can be categorized as series type, parallel type and compound type, of which the compound type hybrid structure is the best solution from the aspects of the fuel efficiency, the additional cost and the expected payback time [6]. In hydraulic systems, energy recovery (ER) is another energy saving method which can be realized by using hydraulic or electrical energy storage devices. The hydraulic approach is to convert the recoverable energy to hydraulic form, store it in a hydraulic accumulator and release it when there are requirements [7–9]. However, it needs additional components such as hydraulic pump/motors or transformers to reuse the recovered energy. The electrical approach, which converts the recoverable energy to electrical form, is more suitable for HHEs, for the hybrid power systems provide electrical energy storage devices such as batteries or super capacitors and the recovered energy can be delivered directly to
⁎ Corresponding author. Tel.: +86 571 87951314 8108; fax: +86 571 87951941. E-mail address: [email protected] (T. Wang).
any electrical actuators. In general, there are two kinds of ER systems in HHEs, including the braking kinetic energy of swings and the gravitational potential energy of booms. The former is similar to the energy regeneration system in hybrid or electrical vehicles which have been studied widely, so this paper concentrates on the latter. A potential energy recovery system for hydraulic forklift trucks was developed and verified in [10]. In comparison to the conventional machines, the efficiency was increased from 56% to 74% at high velocity and from 39% to 69% at low velocity, but the boom response was more oscillatory. In [11], a similar system was studied and the efficiency evaluation of every component is carried out by theoretical analysis and experimental tests. It is reported that the maximum recovery efficiency was 66.2% and improvements could still be achieved. Researches have also been carried out on the HHEs. In [12], additional hydraulic motor-generator was installed in the return oil line of the boom circuit to recover the potential energy, for most hydraulic pumps could not be used in motoring mode. The results of experiment and simulation showed that the ER scheme was feasible in HHEs. Two different ER configurations were compared in [13]. The simulation results showed that the configuration with a hydraulic accumulator had better control performance; however, it was much more complex and its efficiency was only 41% due to the extra link of energy conversion in the accumulator. This paper studies the boom control performance of the HHE with a potential ER system where an electrical generator is driven by a hydraulic motor in the return oil line. The control characteristics of the novel system are analyzed based on the dynamic model. In order to improve the comprehensive performance, solutions including composite control strategy, load torque observation, and leakage flow compensation are
0926-5805/$ – see front matter. Crown Copyright © 2012 Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.autcon.2012.11.034
162
T. Wang et al. / Automation in Construction 30 (2013) 161–169
proposed. The effectiveness of the proposed control scheme is verified by simulations and experiments. The paper is organized as follows. Section 2 presents the structure and parameters of the potential ER system and the experimental platform. Mathematical modeling and analysis are demonstrated in Section 3. Section 4 concentrates on the controller design with the assist of simulations in the MATLAB environment. Section 5 provides the experimental results. Finally, conclusions are drawn in Section 6.
2. Structure and parameters 2.1. System structure The configuration of the compound type HHE with a potential ER system is shown in Fig. 1. It can be seen that the hydraulic pump is driven by the engine and the electrical machine together. The electrical machine behaves as a motor when the load is heavy and a generator when the load is light. By this way, the working condition of the engine is improved and the fuel consumption is reduced. The conventional hydraulic swing motor is replaced with the electrical motor which can be used to recover the braking kinetic energy. The capacitor's state of
charge (SOC) is restrained by a dynamic-working-point control developed in [2]. The boom potential ER system is composed of a hydraulic motor, an electrical generator and a throttle valve. The gravitational potential energy can be converted to the electrical form and stored in the super capacitor when the boom is being lowered down, instead of being dissipated totally in the conventional throttle governing mode. In the proposed mode, the boom velocity is adjusted by controlling the rotational speed of the generator as well as the hydraulic motor. It should be noticed that the throttle valve is necessary in the system, for the oil line should be cut off when the boom is lifted up. Furthermore, the valve can be used to improve the dynamic performance at starting and the low velocity stage, as described in Section 4. Generally, the energy conversion efficiency and the actuator control performance are two important evaluation indexes of the ER system. The total efficiency is mainly determined by the efficiencies of the recovering components and the hydraulic circuits. The control performance indicates the actuator responses to the movement commands which are usually conveyed through joysticks by drivers. Due to the introduction of the ER system, the operation mode becomes different from the conventional mode. Therefore, the velocity governing performance is also changed and should be reconsidered.
Fig. 1. Configuration of the compound HHEs with potential ER.
T. Wang et al. / Automation in Construction 30 (2013) 161–169
2.2. Experimental platform
163
Table 1 Parameters of the experimental platform.
As shown in Fig. 2, an experimental platform is established to study the potential ER system for HHEs. The boom mechanism of a seven-ton hydraulic excavator is employed. The type selection of the key components of the ER system should follow the principles of high efficiency, good control performance and high power density. Thus we apply an axial piston hydraulic motor and a permanent magnet synchronous generator to the platform. The generator is connected with an insulated gate bipolar transistor (IGBT) rectifier. An encoder is used to detect the rotor position and speed information for close-loop control. A DS1104 digital signal processing card is used as the controller where control algorithms are carried out. There are some additional instruments such as boom displacement sensor, torque sensor, pressure sensor and flow meter which are employed to detect the corresponding variables in real time, and these variables are not used for feedback control. The main parameters of the experimental platform are presented in Table 1.
Component
Parameter
Value
Boom cylinder
Rod diameter (mm) Piston diameter (mm) Stroke length (m) Rated flow (L/min) Displacement (mL) Maximum speed (rpm) Rated torque (Nm) Maximum torque (Nm) Maximum current (A) Maximum voltage (V) Capacity (F) Rated voltage (V)
65 115 0.75 100 55 2100 100 160 200 600 6.25 400
Proportional directional valve Hydraulic motor Electrical generator
Rectifier Super capacitor
chambers of the cylinder, A0 and A1 are the corresponding working areas, f is the coulomb friction and Bc is the coefficient of viscous friction. The flow equation of the rodless chamber of the cylinder is derived as
3. Mathematical modeling The first step is to investigate the control characteristics of the ER process with mathematical tools. When the boom is being lowered down, the dynamic equation of the hydraulic cylinder can be given as mc
dvc ¼ F g þ P 0 A0 −P 1 A1 −f −Bc vc dt
ð1Þ
where mc is the equivalent mass of load, vc is the cylinder velocity, Fg is the load force induced by gravity, P0 and P1 are the pressures in the two
V 1 dP 1 ¼ A1 vc −Q v −C 1 P 1 βe dt
ð2Þ
where V1 is the volume of the chamber, βe is the effective bulk modulus of hydraulic oil, C1 is leakage coefficient and Qv is the valve flow which can be expressed as sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2ðP 1 −P 2 Þ Q v ¼ C d Wxv ρ
Fig. 2. Schematic of the experimental platform.
ð3Þ
164
T. Wang et al. / Automation in Construction 30 (2013) 161–169
coefficient. The dynamics of the motor and the generator can be given as Jt
dωm P 2 Dm ¼ η −T e −Bm ωm dt 2π m
ð8Þ
where Jt is the total moment of inertia, ηm is the mechanical efficiency, Te is the electromagnetic torque of the generator and Bm is the coefficient of viscous friction. The permanent magnet generator is a non-salient pole machine, so Te can be expressed as [14] Te ¼
3 pψ i 2 f q
ð9Þ
where p is the number of pole pairs, ψf is the permanent magnet flux linkage and iq is the q-axis current. The voltage equations of the generator in the dq-coordinate are written as Fig. 3. Leakage coefficient of the hydraulic motor.
where Cd is the flow coefficient, W is the area gradient of the valve orifice, xv is the spool displacement, P2 is the pressure in the output port of the valve and ρ is the density of hydraulic oil. The dynamic equation of the spool is obtained by
F m ¼ mv
d2 xv dx þ Bv v þ K v xv dt dt 2
ð4Þ
where Fm is the output force of the proportional solenoid, mv is the mass of the spool, Bv is the coefficient of viscous friction and Kv is the stiffness of spring. The value of Fm is almost linear to the coil current iv and can be expressed as F m ¼ K f iv
ð5Þ
where Kf is the force-current gain. The current dynamics in the solenoid coil is derived as uv ¼ Lv
div dx þ Rv iv þ K b v dt dt
ð6Þ
where uv is the output voltage of the amplifier and is proportional to the control signal, Lv is the inductance and Rv is the resistance of the coil, Kb is the coefficient of the back electromotive force. The flow equation of the inlet chamber of the hydraulic motor is obtained by V 2 dP 2 ω D ¼ Q v − m m −C 2 P 2 βe dt 2π
ð7Þ
where V2 is the volume of the chamber, Dm is the displacement, ωm is the angular speed of the hydraulic motor, and C2 is leakage
ud ¼ −Rs id −Ls
did þ pωm Ls iq dt
ð10Þ
uq ¼ −Rs iq −Ls
diq −pωm Ls id þ pωm ψf dt
ð11Þ
where ud and uq are the output voltages of the vector-controlled rectifier, id is the d-axis current, Rs is the resistance and Ls is the inductance of the winding. According to the equations above, it can be seen that the cylinder velocity is controlled by the opening of the directional valve and the rotational speed of the generator together. A direct strategy is to open the valve fully no matter how much the control command is so as to minimize the losses dissipated in the throttle, and then adjust Te to govern the rotational speed of the generator as well as the cylinder velocity accordingly. However, this strategy is prone to be impactive and oscillatory when the spool moves to the maximum displacement suddenly at starting. In addition, several points should be taken into consideration in this control system. Firstly, the performance of slight operation becomes worse, for the hydraulic motor usually does not work well at the low speed stage. Secondly, the inherent leakage flow of the hydraulic motor affects the anti-disturbance ability and the accuracy of the system. Fig. 3 shows the tested leakage coefficient under various speed of the hydraulic motor. Thirdly, the control link from the generator to the cylinder is an open loop, so the system stiffness is greatly influenced by the close-loop speed stiffness of the generator. Therefore, in order to improve the system performance, the controller should be particularly designed based on the system characteristics. 4. Control system design 4.1. Configuration of control system Fig. 4 shows the block diagram of the control system where uc is the input velocity command, Sc is the SOC of the super capacitor, xv⁎ is the input opening signal of the proportional directional valve, ωm⁎
Fig. 4. Block diagram of the velocity control system.
T. Wang et al. / Automation in Construction 30 (2013) 161–169
165
4.2. Composite control strategy Above all, the system controller should satisfy the conditions that Sc and ωm are no more than their maximum allowable values. Otherwise, the generator should be disabled to avoid damage. Under these conditions, Fig. 5 shows the output signals of the direct strategy and the composite strategy separately. In the direct strategy, the directional valve is fully opened and the generator is enabled to govern the boom cylinder from the starting. In the composite strategy, there are three stages where the valve and generator are controlled with different modes. The principles of the composite strategy are illustrated as follows. Firstly, the recoverable energy is too little to recover and the performance of the hydraulic motor is poor at the low speed stage, so it is better to use the proportional directional valve to control the cylinder and disable the generator at the same time. At this stage, the control signal of the valve is given by
xv ¼ K q uc
ð12Þ
where Kq is the flow gain. Secondly, the directional valve should be opened fully at the high speed stage to minimize the pressure drop on the valve and recover as much energy as possible. Simultaneously the generator is enabled and controlled by the speed signal as
ωm ¼
Fig. 5. Output signals. (a) Direct strategy. (b) Composite strategy.
is the input speed signal and g is the enable signal of the generator. Herein the directional valve and the generator are controlled by a certain strategy where the governing modes vary with different speed stages as described in Section 4.2. The values of Sc and ωm are used as input conditions of the algorithm. Three PI controllers are designed to regulate the currents in Eqs. (6), (10) and (11). The speed of the generator is controlled by a PID controller with load estimation and leakage flow compensation as described in Section 4.3. Since the electrical time constant is generally much smaller than the mechanical time constant, the electrical dynamics have little influence on the dynamic performance of the system and can be regarded as proportional components in the design of the speed controller.
2πvc A1 Dm
ð13Þ
Thirdly, there should be a transition stage between the low and the high speed stage, otherwise impacts and oscillations will be brought in when control signals are sent to the spool at the switch point. The speed control signal of the generator is the same as Eq. (13) at the transition stage. It can be seen that the proportional valve and the generator are governed under different modes according to the values of the input commands in the developed strategy. By using the staged method, the proportional valve is gradually opened and the inlet pressure of the hydraulic motor is gradually increased so that impacts can be eliminated. When slight operation is required, the boom movements are controlled smoothly by the valve and the ER system is disabled. When the boom is being operated fast, the generator is used to govern the cylinder velocity and realize energy recovery. 4.3. Rotational speed governing of the generator The rotational speed control diagram of the generator is shown in Fig. 6. As a simple and reliable method, the PID controller is employed in the main loop. The output of the PID controller is given by t
uðt Þ ¼ K p eðt Þ þ K i ∫0 eðτÞdτ þ K d
Fig. 6. Speed control diagram of the generator.
deðt Þ dt
ð14Þ
166
T. Wang et al. / Automation in Construction 30 (2013) 161–169
where e(t) is the error of angular speed, Kp is the proportional coefficient, Ki is the integral coefficient and Kd is the derivative coefficient. The three coefficients can be obtained with Ziegler–Nichols tuning method. It should be ensured that the generator works in generating mode rather than motoring mode which may result in suction phenomenon. Therefore, the q-axis current output is limited by the saturation of which the upper limit is set to the value corresponding to the maximum electromagnetic torque and the lower limit is set to zero. The system performance is prone to be influenced by the pressure fluctuation in the inlet chamber of the hydraulic motor. In order to reduce the influence and increase the speed stiffness of the generator, a load feedforward loop is designed by using a pressure observer as shown in Fig. 6, instead of using a pressure sensor. According to Eq. (8), the inlet pressure can be estimated by D dωm P^ 2 ¼ m T e þ J t þ Bm ωm GðsÞ 2π dt
where the leakage coefficient can be obtained by the lookup table as shown in Fig. 3. Then the leakage flow is compensated by the following expression
ω ˜ m ¼ ωm −
2πQ l Dm
It can be observed from Eq. (17) that the compensated value will become minus when the speed command is small enough. It means that the generator should be operated as a motor to satisfy the low speed command. This phenomenon is energy-consuming and obviously diseconomical. Fortunately, it can be avoided by using the composite
ð15Þ
where Te can be obtained from the control command, ωm can be obtained by the speed feedback and G(s) is a low-pass filter which is used to depress the noise brought in by the derivation of ωm. Since the load is observed, the leakage flow of the hydraulic motor can be calculated as Q l ¼ C ðωm ÞP^ 2
ð16Þ
Fig. 7. Simulation results under different strategies.
ð17Þ
Fig. 8. Experimental results with potential ER (vc = 0.13 m/s).
T. Wang et al. / Automation in Construction 30 (2013) 161–169
167
of the two different control strategies before experiments. Fig. 7 shows the simulation results. It can be seen that the system with the direct strategy has considerable oscillation at starting. However, by using the proposed composite strategy, the performance can be greatly improved. 5. Experimental study 5.1. Control performance Fig. 8 shows the experimental results with potential ER and Fig. 9 provides the results of throttle governing for comparison. The results indicate that the boom control performance under the proposed
Fig. 9. Experimental results using throttle governing (vc = 0.13 m/s).
strategy described above because the generator is disabled at the low speed stage.
4.4. Simulations Based on the mathematical model, simulations of the control system are carried out in the MATLAB environment to verify the performances
Fig. 10. Experimental results at low velocity (vc = 0.013 m/s).
Fig. 11. Comparison of systems without and with load observation.
168
T. Wang et al. / Automation in Construction 30 (2013) 161–169
strategy is close to the performance of throttle governing although there is a little overshoot. The voltage of the super capacitor which represents the SOC is obviously increased in the ER process. Fig. 10 presents the experimental results at low boom velocity. It can be seen that there are ripples when the boom is controlled by the ER system. So it is proper to control the boom by using a valve when the velocity command is very small as demonstrated in Section 4. Fig. 11 gives the boom velocities of the systems without and with load torque observation. As shown in Fig. 11, the pressures in the rodless chamber of the boom cylinder are varied from 11 MPa to 9 MPa by adjusting the relief valve to simulate the load variation. It can be seen that the velocity fluctuation with load observation is depressed by approximate 50%. The results show that the software method is effective in improving the system stiffness without additional cost of sensors. Fig. 12 gives the boom velocities of the system without and with leakage compensation. It can be found that the accuracy of the compensated system is improved. In fact, the leakage flow of the hydraulic motor is various with the load pressure, so the anti-disturbance capability of the control system can also be enhanced with the compensation. 5.2. Efficiency evaluation A series of experiments with different boom velocities are carried out to investigate the efficiency of the potential ER system. Considering from the aspects of power and energy respectively, the maximum and the average energy conversion efficiencies from the rodless chamber of the boom cylinder to the super capacitor are given as ηec
max
ηec
ave
U c Ic ¼ max Pc Q c ¼
∫U c Ic ∫P c Q c
ηem
max
ave
U c Ic ¼ max Pm Q m ¼
∫U c I c ∫P m Q m
Boom velocity (m/s)
ηem_ave
ηec_ave
ηem_max
ηec_max
0.013 0.025 0.05 0.07 0.10 0.13
0.000 0.549 0.649 0.693 0.662 0.587
0.000 0.467 0.608 0.645 0.542 0.419
0.000 0.756 0.762 0.773 0.804 0.811
0.000 0.635 0.713 0.714 0.657 0.583
where Pm and Qm are the pressure and the flow rate in the inlet chamber of the hydraulic motor. The experimental results are shown in Table 2. The efficiencies are zeros when the boom velocity is very low and the generator is disabled. Relatively high efficiencies can be obtained at the middle velocity stage. It can be deduced that a part of potential energy is dissipated in the directional valve by analyzing the data. Therefore, better results can be achieved by properly selecting the directional valve. The value of ηem_max, which represents the transient energy conversion efficiency of the hydraulic motor and generator, is improved with the increase of the boom velocity, for the generator usually obtains the maximum efficiency when the practical power is equal to its rated power. In general, the potential ER efficiency is acceptable in this experimental platform. 6. Conclusion
ð18Þ
ð19Þ
where Pc and Qc are the pressure and the flow rate in the rodless chamber of the boom cylinder, Uc and Ic is the voltage and the charging current of the super capacitor. The maximum and the average energy conversion efficiencies from the inlet of the hydraulic motor to the super capacitor are given as ηem
Table 2 Test of energy recovery efficiency.
ð20Þ
ð21Þ
The boom control performance of the HHE with a potential ER system is studied in this paper. The characteristics of the control system are analyzed by modeling and simulation. Solutions including composite control strategy, load torque observation and leakage flow compensation are proposed to improve the dynamic performance of the system and maintain relatively high recovery efficiency. The experimental results show that the proposed control scheme can recover the potential energy of the boom effectively with acceptable control performance. The further work is to research the power management when the prototypical potential ER system is applied to a working HHE model. Acknowledgment The authors gratefully acknowledge the financial support of the National Natural Science Foundation of China (Grant No. 50875233) and the National High Technology Research and Development Program of China (Grant No. 2010AA044401). References
Fig. 12. Comparison of systems without and with leakage compensation.
[1] Q.F. Wang, Y.T. Zhang, Q. Xiao, Evaluation for energy saving effect and simulation research on energy saving of hydraulic system in hybrid construction machinery, Chinese Journal of Mechanical Engineering 41 (12) (2005) 135–140. [2] Q. Xiao, Q.F. Wang, Y.T. Zhang, Control strategies of power system in hybrid hydraulic excavator, Automation in Construction 17 (4) (2008) 361–367. [3] M. Ochiai, S. Rye, Hybrid in construction machinery, in: Proceedings of the 7th JFPS International Symposium on Fluid Power, Toyama, 2008, pp. 41–44. [4] D.Y. Wang, C. Guan, S.X. Pan, M.J. Zhang, X. Lin, Performance analysis of hydraulic excavator powertrain hybridization, Automation in Construction 18 (3) (2009) 249–257. [5] J. Choi, H. Kim, S. Yu, K. Yi, Development of integrated controller for a compound hybrid excavator, Journal of Mechanical Science and Technology 25 (6) (2011) 1557–1563. [6] T.-S. Kwon, S.-W. Lee, S.-K. Sul, C.-G. Park, N.-I. Kim, B.-I. Kang, et al., Power control algorithm for hybrid excavator with supercapacitor, IEEE Transactions on Industry Applications 46 (4) (2010) 1447–1455. [7] X.G. Liang, T. Virvalo, Development and research of an energy saving drive in a hydraulic crane, in: The 7th Scandinavian International Conference on Fluid Power, Sweden, 2001, pp. 151–161. [8] T. Virvalo, W. Sun, Improving energy utilization in hydraulic booms - what it is all about, The 6th International Conference on Fluid Power Transmission and Control, Hangzhou, 2005.
T. Wang et al. / Automation in Construction 30 (2013) 161–169 [9] J.D. Van de Ven, M.W. Olson, P.Y. Li, Development of a hydro-mechanical hydraulic hybrid drive train with independent wheel torque control for an urban passenger vehicle, in: Proceedings of the International Fluid Power Exposition, Las Vegas, Nevada, 2008. [10] T.O. Andersen, M.R. Hansen, H.C. Pedersen, F. Conrad, Regeneration of potential energy in hydraulic forklift trucks, in: The 6th International Conference on Fluid Power Transmission and Control, Hangzhou, 2005, pp. 302–306. [11] T. Minav, P. Immonen, L. Laurila, V. Vtorov, J. Pyrhonen, M. Niemela, Electric energy recovery system for a hydraulic forklift – theoretical and experimental evaluation, IET Electric Power Applications 5 (4) (2011) 377–385.
169
[12] Y.T. Zhang, Q.F. Wang, Q. Xiao, Simulation and experimental research on energy regeneration with hydraulic motor for hybrid drive excavator, Chinese Journal of Mechanical Engineering 43 (8) (2007) 218–223. [13] T.L. Lin, Q.F. Wang, B.Z. Hu, W. Gong, Research on the energy regeneration systems for hybrid hydraulic excavators, Automation in Construction 19 (8) (2010) 1016–1026. [14] M. Chinchilla, S. Arnaltes, J.C. Burgos, Control of permanent-magnet generators applied to variable-speed wind-energy systems connected to the grid, IEEE Transactions on Energy Conversion 21 (1) (2006) 130–135.