A Stochastic Predictive Energy Management Strategy for Plug-in Hybrid Electric Vehicles Based on Fast Rolling Optimization

• Published in: IEEE transactions on industrial electronics (1982) Vol. 67; no. 11; pp. 9659 - 9670
• Main Authors: Yang, Chao, You, Sixiong, Wang, Weida, Li, Liang, Xiang, Changle
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.11.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Batteries; Chain scission; Computer simulation; Computing time; Consumption; Dynamic programming; Electric vehicles; Energy consumption; Energy management; Energy management strategy; Engines; fast optimization; Fuel economy; Fuels; Hybrid electric vehicles; Markov chains; Mechanical power transmission; model predictive control (MPC); Optimization; plug-in hybrid electric vehicles (PHEV); Powertrain; Predictive control; Run time (computers); stochastic driving behavior; Stochastic models; Strategy; Torque
• ISSN: 0278-0046; 1557-9948
• DOI: 10.1109/TIE.2019.2955398

This article proposes a stochastic predictive energy management strategy based on fast rolling optimization for plug-in hybrid electric vehicles (PHEVs). First, combined with a large number of real-world driving cycle data, the stochastic driving behaviors are modeled as probability transition matrices of vehicle demand torque based on Markov chains. Secondly, to solve the torque split problem in parallel hybrid powertrain, the stochastic model predictive control (SMPC) framework is built. Thirdly, the continuation/generalized minimum residual algorithm is employed to execute the fast rolling optimization. The effectiveness of the proposed strategy is validated in both simulations and test bench, and its performance is compared with the SMPC by dynamic programming (DP) optimization and the equivalent minimum fuel consumption strategy (ECMS). Simulation results show that under real-world driving cycle, PHEV using the proposed strategy could obtain 4.8% energy consumption reduction comparing with that uses ECMS. In terms of computational time, the proposed strategy dramatically reduces the running time comparing with that of SMPC by DP optimization. Furthermore, the similar results can be obtained in the experiment. Under real-world driving cycle, 4.6% fuel economy improvement is obtained using the proposed strategy compared with that using ECMS, which clearly shows that the proposed strategy is effective.