An Energy Management Method of Electric Vehicles Based on Stochastic Model Predictive Control

Hybrid electric vehicles (HEVs) are a feasible way to solve the problem of environmental pollution and energy shortage. Aiming at the problem of error of energy management strategy based on stochastic dynamic programming in driver demand torque prediction model, the application of stochastic model predictive control with feedback correction and rolling optimization in energy management was studied. After a brief analysis of the optimization problem, the uncertainty of vehicle operating conditions is expressed by a one-step Markov model of driver's demand torque. Then, the state space equation of the predictive model of single-axle parallel hybrid electric vehicle in stochastic model predictive control is given. Then, dynamic programming method and multi-step instant differential learning method are applied in rolling optimization to reduce the computational complexity and improve the real-time performance of the algorithm. Feedback correction in stochastic model predictive control was used to reduce the predictive error caused by probabilistic predictive model. At the same time, dynamic programming and multi-step instant differential learning were combined to solve the control action in rolling optimization, which reduced the computational complexity of the algorithm. Through simulation, the average energy recovery efficiency of the motor is 94.12%, and the SOC range is 0.56-0.82.