Optimal Design of a Compound Hybrid System consisting of Torque Coupling and Energy Regeneration for Hydraulic Hybrid Excavator *

Optimal Design of a Compound Hybrid System consisting of Torque Coupling and Energy Regeneration for Hydraulic Hybrid Excavator * Yang Xiao, Cheng Guan, Perry Y. Li and Fei Wang Abstract For hydraulic hybrid excavators (HEEs), two recoverable energy sources for power hybridization are from engine optimization and actuator energy regeneration. This paper presents a hydraulic hybrid system that integrates with these two sources using engine torque coupling and boom energy regeneration, thus to further explore the fuel economy potential of HHE. Their different characteristics are analyzed first to perform specialized system design. Then an optimal control problem is formulated and solved for system sizing. The effectiveness of the proposed system and its energy saving performance will be assessed, as well as the contributions of engine torque coupling and boom ERS, respectively. Abstract Hydraulic hybrid excavator, torque coupling, energy regeneration, optimal design, fuel economy I. INTRODUCTION Powertrain hybridization has drawn a lot of researchers attention in recent years, with the aim to achieve higher fuel economy due to the concerns about energy crisis and environment pollution. With the benefit of high power density and low component price, hydraulic hybrid technology has found its way to make contribution in industry applications, especially high energy consuming realms. Much of the focus of hydraulic hybrid research has been on vehicles [-6]; however, there has been a growing interest from academia and industry in hydraulic hybrid excavator [7-9]. There are two recoverable energy sources in a hydraulic excavator, one is from the engine optimization and the other one is from the energy regeneration of hydraulic actuators [-2]. As the engine speed is relatively more stable in an excavator compared to a vehicle, torque coupling which applies a secondary component to output or absorb torque so as to assist the engine, is a proper structure to perform engine optimization for excavators [8]. Meanwhile, the energy recovered from hydraulic actuators is in hydraulic form, which makes it suitable to be integrated into hydraulic hybrid system. However, few researches have been reported on combing these two sources together. In this study, a compound hybrid system that consists of torque coupling and boom Energy Regeneration System (ERS) is designed based on our previous work [8, ]. Then an optimal control problem is formulated and solved for system Yang Xiao, Cheng Guan and Fei Wang are with the Institute of Mechanical Design, Zhejiang University, Hangzhou 327, China (email: xyboy7@hotmail.com, guan @zju.edu.cn, wf_car@63.com) Professor Perry Y. Li is with the Department of Mechanical Engineering, University of Minnesota, Minneapolis, MN 55455, USA (email: lixxx99@umn.edu). *A portion of this work was performed at University of Minnesota. parameterization, as well as for exploring the energy saving potential of engine torque coupling and boom ERS. This paper is organized as following: the compound hybrid system design is presented in Part II. System modelling is given in Part III. Then the optimal control problem formulation, solving and system optimization procedures are introduced in Part IV. The results and discussion are presented in Part V. Conclusions are offered in the last part. II. SYSTEM CONFIGURATION A GC228LC 23 ton excavator is used as the base in this study. A hybrid configuration shown in Figure is proposed, with the key feature of combining power hybridization structure and energy regeneration system. A. Power distribution system The torque coupling structure is applied to perform power hybridization. Through the transfer case, the engine and the secondary component pump/motor together supply the load power on the hydraulic pump. As long as the pressure conditions of hydraulic accumulator are satisfied, the pump/motor could be used to adjust the engine torque output, thus to regulate engine operating points within desired high efficiency region. A mechanical controlled engine is used in this study, which makes it proper to set engine throttle at fixed positons during working cycles because of the slow dynamics of engine throttle control. B. Boom energy regeneration system Besides engine optimization, the proposed system incorporates boom ERS that captures gravitational potential energy of the boom actuator as well. With the use of switch signal C 2, the boom ERS can work under conventional mode and hybrid mode. In conventional mode, C 2 signal is deactivated, and the ERS system equals to the conventional system, where the direction of flow is determined by the multiple directional valve. In energy regeneration mode, the controller detects the handle, and relates C 2 to boom descending signal. During boom descending, active C 2 would change the circuit so that the pressured oil in the lower chamber of boom cylinder would automatically be directed into the accumulator as long as the maximum pressure condition of the accumulator is satisfied. C. Energy storage system The energy storage system plays a crucial role in the integrated system of power hybridization and energy regeneration. With the consideration of reducing the pump/motor size as well as increasing efficiency, the maximum pressure of torque coupling related accumulator should be set high enough so that we can take advantage of the

Accumulator Accumulator 2 Boom ERS P u P u C2 Valve2 Valve C Valve3 Pump/Motor Valve4 Multiple directional valve Handle Engine Pump Pump 2 Handle SOC Controller C C2 DP/M Engine Transfer Case Figure. Compound hydraulic hybrid excavator system. high power density of hydraulic accumulator. On the other hand, for boom ERS, the pressure of the lower chamber stays in relative lower range in conventional mode. And when the regeneration mode is activated, the related accumulator pressure acts as an equivalent back pressure for boom descending during energy regeneration. High accumulator pressure would cause excessive equivalent back pressure for boom descending as well as high system power demand which would eventually devastate the drive feelings. This brings up a lower pressure demand for the boom ERS related accumulator compared to engine torque coupling. Therefore, a structure of two separate accumulators is applied as the energy storage system. The high pressure accumulator is used to expand the operating pressure range so as to make better use of the auxiliary power source, while the low pressure accumulator 2 is used as regulated equivalent back pressure for boom ERS. Then the directional valve Valve is used to determine which accumulator is used to supply the power distribution system. III. SYSTEM MODELLING A. Conventional s Figure 2 gives a working cycle of a conventional excavator over typical excavation tasks in experiments. Because of the periodic excavation movements, the working cycle shows periodic feature, with a period time around 5 seconds. Meanwhile, the speed changing range is within 2 rpm, relative narrower compared to regular vehicles that have drastic speed change. The engine operating points on engine map are shown in Figure 3. Given the characteristics of mechanical controlled engine, the engine speed varies along a certain torque-speed changing curve under fixed engine Pressure (MPa) Pump Torque (Nm) Pump Power (kw) 4 2 5 55 6 65 7 75 8 85 9 95 2 22 2 2 9 5 55 6 65 7 75 8 85 9 95 2 5 5 55 6 65 7 75 8 85 9 95 2 5 5 5 55 6 65 7 75 8 85 9 95 2 Time (s) Figure 2. Power demand of conventional system. Pump Pump2 throttle. As engine speed remains stable, engine torque varies according to load, leaving the engine operating points scattered and away from high fuel consumption rate area. Thus, torque coupling which use pump/motor to adjust the engine output torque, is a proper structure to perform power hybridization task. Our goal is then to use torque coupling to move these points to desired high efficiency area. B. Engine The engine dynamics can be summarized as follows:

Torque (Nm) 6 5 4 3 2 24 26 24 24 26 28 3 35 39 43 24 26 28 3 35 39 43 48 4 6 8 2 22 Volumetric Efficiency.8.6.4.2 6 4 2 Rotate Speed(rpm) - Displacement Ratio.9.8.7.6.5.4.3.2. Figure 3. Engine behavior of the conevntional system over typical. J ω = T T T () eng eng eng pm load where J eng is the rotational inertia, ω eee is the engine speed, T eee is the output torque of engine, T pp is output torque of pump/motor, and T llll is the torque load. As for hydraulic excavator system, the actuation of working devices are supplied by the main pump, so it is proper to define the torque load by torque demand on main pump in Figure 2. Experiment data interpolation is used for mapping the engine fuel consumption rate, and the total fuel consumption is: m = m ( N, T, ω, u ) dt (2) fuel fuel load eng eng th where m ffff is the total fuel consumption, m ffff is the engine fuel consumption rate, N llll is the system load power on main pump, and u th is the engine throttle, which is set as constant in this study. The transfer case is used to realize torque coupling between engine and secondary component, and the gear ratio is set to in this study: ωeng = ωpm = ω (3) pump where ω eee is engine speed, ω pp is the pump/motor speed, and ω pppp is the main pump speed. C. Pump/Motor The size of pump/motor is to be optimized in this study. Thus, a set of scalable torque and flow characteristics maps based on a typical 28 mm 3 variable displacement bent-axis pump/motor are used. As the torque and flow are assumed to scale linearly with displacement variation, the torque and flow efficiencies become invariant with respect to the torque and flow. Figure 4 shows the torque and volumetric efficiency map under constant pressure 5 MPa. Both efficiencies remain relative high at some operating conditions, while they drop significantly at low displacement cases. And the torque and flow equations can be calculated as: Ppm Dpm T = x η P x 2p (, ω, ) pm pm T pm pm pm (,, ) Q = x D ω η P ω x pm pm pm pm Q pm pm pm ( ) sign x pm ( ) sign x pm (4) (5) Mechanical Efficiency.8.6.4.2 6 4 2 Rotate Speed(rpm) Figure 4. where P pp is the pressure difference between pump/motor, x pp is the displacement ratio, η T is the torque efficiency, and η Q is the volume efficiency. Positive x pp means pump mode and negative x pp means motor mode. D. Accumulator To model the energy storage system, each accumulator model is first formulated based on Boyle s Law: E acc Torque and vlumetric efficiency of the pump/motor. P V = P V = const (6) n - i n i n V i n PV P = PdV = n P V i where P is the pre-charging pressure, V is the pre-charge air volume which is also taken as the rated accumulator volume, E acc is the hydraulic energy contained in accumulator and n is the polytrophic exponent (n =.4 for nitrogen). Then with the use of directional valve Valve, the equivalent pressure and flow that equals to the pressure and flow of pump/motor are expressed by following equations: ( ) acc _ eq pm d acc d acc2 (7) P = P = x P + x P (8) Displacement Ratio.9.8.7.6.5.4.3.2.

( ) Q = Q = x Q + x Q (9) acc _ eq pm d acc d acc2 where P acc, P acc2 are the pressure of accumulators, Q acc, Q acc are the flow rate of the accumulators, and the value of directional valve x d equals to either indicating the left position, or indicating the right position. E. Boom ERS In conventional system, the throttle in the multiple directional valves that connected to the lower chamber of boom cylinder is used to add back pressure for boom descending movement. However, the hydraulic oil in the lower chamber is lead to the accumulator in boom ERS rather than to the original path, making accumulator pressure the equivalent back pressure instead. Thus, pressure constraints are required to regulate the pressure of accumulator 2. The boom cylinder is described as: m v = P A + F P A () b b upper upper m lower lower Qrec = Alower v () b where m b is the mass of boom cylinder, v b is the speed of boom cylinder, P uuuuu is the upper chamber pressure, P lllll is the lower chamber pressure, A uuuuu and A lllll are sectional area of upper and lower chambers, F m is the equivalent boom cylinder weight, Q rrr is the recovered flow from boom ERS. The lower limit of the accumulator pressure is chosen in the static situation that P uuuuu is the threshold of the pressure controlled valve Valve2 in Figure to activate the boom ERS, i.e..5 MPa. The upper limit of this range is chosen based on the specific characteristics of constant-power negative-flux system used in this study. The upper limit of P upper is chosen as the threshold P thres of constant-power. Then the constraint is defined as: P acc2 F +.5 A F + P A, Alower Alower m upper m thres upper (2) For simplicity, one assumption is made here that the hydraulic oil in boom lower chamber would all go into the accumulator as long as the pressure constraint is satisfied. IV. SYSTEM OPTIMIZATION A. Optimal control formulation and solving Optimal control is used to evaluate the energy saving performance of the proposed hybrid configuration. The definition of optimality is to minimize engine fuel consumption over the whole. Using the typical conventional given above as the system demand, a static optimization problem can be formulated as: ( w ) m = min m N, T, dt (3) * fuel fuel load eng eng weng, Teng where m ffff is the optimal fuel consumption. T eee, ω eee are chosen as the control variables of energy management. In torque coupling structure, once engine output is determined, pump/motor output is determined as well. Terminal constraints on the accumulators are needed to constrain the accumulator behavior, and are formulated as: ( ) ( ) SOC t SOC t P Q dt f = acc acc = ( ) ( ) 2 f 2 acc2 acc2 rec (4) SOC t SOC t = P ( Q + Q ) dt = (5) The constraints are set due to the self-sustaining consideration of returning the state of charge (SOC) to their initial values at the end of the specific, meaning the total discharging equal charging energy, and total energy output is zero. Considering the energy regeneration effect, the constraint for accumulator 2 is modified with the recovered flow Q rrr, which indicates that the charging energy consists of energy from engine torque coupling and boom ERS. Deterministic dynamic programming is generally used to solve the constrained optimization problem of the hybrid configuration over a given. However, with the assumption of unrestricted size of the accumulator, which equals to constant accumulator pressure, the computation can be significantly simplified. This analysis method can not only provide rapid optimal solution for a prescribed, but also give insight into the energy management control. [-4] Based on this assumption, a Lagrange Multiplier (LM) method can be used to solve the optimal control problem over given. By adding the terminal constraints into equation 3 with λ, λ 2, the problem is re-formulated as: l, l2 weng, Teng ( ( w ) max min m N, T,? fuel load eng eng ( l acc acc l2 acc2( acc2 rec )) ) + P Q + P Q + Q lhv dt (6) where lhv is the lower heating value of diesel. Then the constraint optimal problem can be solved by the Min-max problem in equation 6. While the minimization ensures the optimality over different λ, the maximization ensures the satisfaction of the terminal constraints as long as there is solution to the problem. And the optimal λ, λ 2 can then be calculated as constants that summarize the whole. Using the optimal λ, λ 2, the control law can be derived as: ( w ) * * [ weng, Teng ] = arg min ( mfuel Nload, Teng, eng weng, Teng * * ( l acc acc l2 acc2( acc2 rec )) + P Q + P Q + Q lhv) dt (7) B. System Sizing As the system structure is determined, system optimization becomes a process of optimize fuel economy under different system parameter choices. The overall optimization process can be summarized as: ) Iterate the pump/motor size from defined range 2) Solve the optimal control problem in (Equation 6) 3) Check the constraint satisfaction. If fails, go to step 5 4) Evaluate the generated fuel economy performance

Fuel consumption/kg Fuel consumption/kg 6.36 6.35 6.34 6.33 6.32 6.3 Minimum(4mm 3, 6.3kg) 6.3 8 2 4 6 8 2 Pump/motor size/mm 3 5.55 5.5 5.45 5.4 5.35 Minimum (3mm 3, 5.35kg) Torque (Nm) 6 5 4 3 2 6 26 2424 24 3 4 5 6 7 8 9 2 2 22 26 26 2424 24 26 28 28 3 3 35 35 39 39 4343 48 8 2 4 6 8 2 Pump/motor size/mm 3 5 24 24 Figure 5. Fuel consumption over pump/motor size: ) torque coupling system; 2) compound hybrid system. 5) Iterate step 2, 3, 4 with new pump/motor size, to minimize the fuel consumption 6) Evaluate the optimized design s fuel economy and the contribution of engine torque coupling and boom ERS V. RESULTS AND DISCUSSIONS To better study the respective effect of engine torque coupling and boom ERS, a comparison system named torque coupling system is also optimized and analyzed. Torque coupling system refers to the proposed compound hybrid system with inactive boom ERS, and is used to show the performance of engine torque coupling alone. And the different performance between torque coupling system and compound hybrid system indicates the effect of boom ERS. A. Optimization results The preset constant pressure of accumulator and accumulator 2 are 2 MPa and 8 MPa, as set by the pre-charge pressure. Figure 5 shows the fuel consumption over pump/motor size, and the optimal size is chosen as the minimum. If the pump/motor size is smaller than the optimal, the effect of power assistance would be limited. However, if the size is larger than the optimal, the fuel consumption would increase since same displacement demand would cause smaller displacement ratio, which would drop the pump/motor efficiency as well as the overall energy saving effect. Figure 6 shows the optimal engine behavior of the compound hybrid system over the. One obvious difference compared to the conventional system in Figure 3, is the distribution of engine operating points. Along the speed control curve, the engine mainly operates within two areas. One is the most efficiency area where the fuel consumption rate is the best among all engine choices, while the other one is the least energy loss area, where the engine output is relative low and the total amount of energy loss is the least. Further insight into the optimal power distribution between engine, pump/motor and boom energy regeneration Torque (Nm) 4 3 2 26 3 4 5 6 7 8 9 2 2 22 Figure 6. Engine behavior over typical : ) torque coupling system; 2) compound hybrid system. Power (kw) Power (kw) 5-5 5-5 Engine Power P/M Power are shown in Figure 7. Positive pump/motor power indicates motor mode, and negative pump/motor indicates pump mode. Positive boom ERS indicates energy charging mode. For the torque coupling system, the optimal control law operates the engine either at the most efficient points or at the least loss 26 28 28 3 3 35 35 39 39 4343 Figure 7. Power distribution of the compound hybrid system: ) torque coupling system; 2) compound hybrid system. 48 2 4 6 8 2 Load Power (kw) Engine Power P/M Power Boom ERS 2 4 6 8 2 Load Power (kw)

TABLE I. Engine Energy Output (MJ) Fuel Consumption (kg) Engine efficiency (%) Fuel efficiency improvement (%) ENERGY SAVING COMPARSION Conventional Torque Compound coupling 84.45 92.8 78.2 7.3 6.3 5.35 3.2 38. 37.83 N/A.24 23.9 points with a clear shifting threshold. And the pump/motor is used to compensate for the difference between engine and load demand. For the compound system with active boom energy regeneration, besides the similar power distribution behavior, there are certain differences when the boom ERS is active. This is because that the extra recovered energy from boom changes the power distribution pattern when minimizing fuel cost, leaving the engine to operate at sub-optimal points. Therefore, analysis on the energy saving effects is needed to check the effect of engine torque coupling and boom ERS. B. Energy saving effect comparison A detailed energy saving effect comparison between engine torque coupling and boom ERS is made, and the results are given in table I. Engine energy output means the total energy outputted by engine throughout the. Compared to the conventional system, torque coupling system consumes more energy to fulfill the power demand. It is due to the efficiency of pump/motor during torque coupling period, and the engine has to output extra energy for the pump/motor energy loss. While for the compound system, although the pump/motor efficiency issue still remains, the recovered energy from boom ERS plays a more dominate role to reduce the total engine energy output to 78.2 MJ. Fuel consumption indicates the total fuel consumed by engine. And the engine efficiency is defined as the ratio of engine energy output and fuel consumption: η eee = E eee m ffff lhv. (8) With the use of torque coupling hybridization, the engine efficiency is improved from 3.2% to38.%. While for the compound system, the engine efficiency is 37.83%, slightly lower than torque coupling system. It can be explained by power distribution differences in Figure 7 when boom ERS is active. Lower operating pressure set for boom energy regeneration would decrease the pump/motor efficiency and bring down the effect of engine optimization in compound hybrid system. However, the engine efficiency remains at the same level as the torque coupling system. The fuel consumption improvement results then not only prove the effectiveness of the proposed compound hybrid system, but also give an indication of the benefits of torque coupling and boom ERS for the overall energy saving effect. As torque coupling system alone can improve the fuel economy by.24%, the compound system which integrates both engine torque coupling and boom ERS can advance the fuel economy to 23.9%. VI. CONCLUSION In this paper, a compound hybrid system which integrates torque coupling and boom energy regeneration is proposed for hydraulic hybrid excavator. Based on the different characteristics of the energy in engine torque coupling and boom ERS, the energy storage system is design with two accumulators, one high pressure and one low pressure. An optimal control problem is then formulated and solved to perform system optimization for the compound hybrid system. The generated optimization results not only show the effectiveness of the proposed compound hybrid system, but also show the influence of engine torque coupling and boom ERS on the overall energy saving effects. ACKNOWLEDGMENT This work is supported by the National High Technology Research and Development Program of China (grant number 2AA444). REFERENCES [] P. Y. Li and F. Mensing, "Optimization and control of a hydro-mechanical transmission based hybrid hydraulic passenger vehicle," in 7th International Fluid Power Conference, Aachen, 2. [2] K. L. Cheong, P. Y. Li, and T. R. Chase, "Optimal Design of Power-Split Transmission for Hydraulic Hybrid Passenger Vehicles," in American Control Conference, San Francisco, CA, USA, 2, pp. 3295-33. [3] D. Zhekang, C. Kai Loon, P. Y. Li, and T. R. Chase, "Fuel economy comparisons of series, parallel and HMT hydraulic hybrid architectures," in American Control Conference (ACC), 23, 23, pp. 5954-5959. [4] D. Zhekang, C. Kai Loon, P. Y. Li, and T. R. Chase, "Energy Management Strategy of a Power-split hydraulic hybrid vehicle based on Lagrange Multiplier and Its Modifications," (under review). [5] Z. Filipi and Y. J. Kim, "Hydraulic Hybrid Propulsion for Heavy Vehicles: Combining the Simulation and Engine-In-the-Loop Techniques to Maximize the Fuel Economy and Emission Benefits," Oil & Gas Science and Technology-Revue D Ifp Energies Nouvelles, vol. 65, pp. 55-78, Jan-Feb 2. [6] L. Jinming and P. Huei, "Modeling and Control of a Power-Split Hybrid Vehicle," Control Systems Technology, IEEE Transactions on, vol. 6, pp. 242-25, 28. [7] R. Hippalgaonkar and M. Ivantysynova, "A Series-Parallel Hydraulic Hybrid Mini-Excavator with Displacement Controlled Actuators," in The 3th Scandinavian International Conference on Fluid Power, SICFP23, Linköping, Sweden, 23. [8] D. Y. Wang, C. Guan, S. X. Pan, M. J. Zhang, and X. Lin, "Performance analysis of hydraulic excavator powertrain hybridization," Automation in Construction, vol. 8, pp. 249-257, May 29. [9] T. Wang and Q. F. Wang, "Modeling and control of a novel hydraulic system with energy regeneration," in Advanced Intelligent Mechatronics (AIM), 22 IEEE/ASME International Conference on, 22, pp. 922-927. [] T. L. Lin, Q. F. Wang, B. Z. Hu, and W. Gong, "Research on the energy regeneration systems for hybrid hydraulic excavators," Automation in Construction, vol. 9, pp. 6-26, 2. [] Y. Xiao, C. Guan, and X. Lai, "Research on the design and control strategy for a flow-coupling-based hydraulic hybrid excavator," Proc IMechE, Part D: Journal of Automobile Engineering, vol. 228, pp. 675-687, December 24. [2] A. Chauvin, A. Sari, A. Hijazi, and E. Bideaux, "Optimal sizing of an energy storage system for a hybrid vehicle applied to an off-road application," in Advanced Intelligent Mechatronics (AIM), 24 IEEE/ASME International Conference on, 24, pp. 775-78.