Abstract:Considering different power requirements of the tractor under various working conditions, the singlemotor powertrain system is usually in low load state under low load working conditions, with low efficiency, resulting in energy waste. To solve this problem, a dualmotor multimode coupling driving system for electric tractors was proposed. Through the coordinated control of two motors and brakes, four driving modes could be realized: motor EM_S independent drives, motor EM_R independent drives, dualmotor coupling drives and dualmotor independent drives. These modes could meet the tractor power demand under various working conditions, improve the load rate of motor, and thus improve the efficiency of the whole vehicle. A parameter matching method was proposed to match the parameters of the two motors and the main transmission ratios of the coupling box according to the dynamic performance indexes of the typical working conditions. In order to realize the efficient operation of DMET, and enhance the adaptability of the energy management strategy to different operating conditions of electric tractors, a realtime energy management strategy based on stochastic dynamic programming + extreme seeking algorithm (SDP_PESA) was proposed. The state feedback control table generated offline by SDP as the control input reference was used to ensure the approximate global optimum. On this basis, the adaptive optimization algorithmPESA was introduced to dynamically search the local maximum value of the system output to compensate for the control input of SDP and generate operating points with lower energy consumption and higher efficiency. The EMS based on SDP_PESA considered the good optimization performance of the global optimization algorithm and the robustness of the instantaneous optimization algorithm comprehensively, and the advantages of the two algorithms were used to achieve more excellent control performance. A DMET simulation model with SDP state feedback control table was established based on Matlab/Simulink, and real operating conditions were used to simulate the energy management strategies based on SDP and SDP_PESA. The simulation results demonstrated that the actual vehicle speed can track the change of the target vehicle speed in real time, and the control strategy can quickly respond to the change of the work load, indicating that the DMET simulation model was efficient and feasible, and can meet the simulation accuracy requirements. Based on the energy management strategy of SDP, the average power consumption of DMET in plowing and transportation conditions were respectively 1.77kW·h/km and 1.17kW·h/km, the driving efficiency was 0.80 and 0.81, respectively. However, after adding the PESA output feedback controller, the driving efficiency was increased by 213% and 197%, and the average power consumption was reduced by 10.17% and 16.2%, respectively, which meant the energy management strategy based on SDP_PESA can effectively increase the operating range of pure electric tractors, and SDP_PESA was fully capable of realtime application.