A multi-objective optimization energy management strategy for power split HEV based on velocity prediction

• Published in: Energy (Oxford) Vol. 238; p. 121714
• Main Authors: Wang, Weida, Guo, Xinghua, Yang, Chao, Zhang, Yuanbo, Zhao, Yulong, Huang, Denggao, Xiang, Changle
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.01.2022; Elsevier BV
• Subjects: Acceleration; Algorithms; Consumption; Cost function; Deceleration; Driving conditions; Electric power distribution; Electric vehicles; Energy management; Energy management strategy; Energy storage; Fuel consumption; Fuel economy; Hybrid electric vehicle; Hybrid electric vehicles; Hybrid energy storage system; Hybrid systems; Model predictive control; Multi-objective optimization; Multiple objective analysis; Neural networks; Optimization; Power consumption; Predictions; Predictive control; Velocity; Model predictive control; Hybrid energy storage system; Multi-objective optimization; Energy management strategy; Hybrid electric vehicle
• ISSN: 0360-5442; 1873-6785
• DOI: 10.1016/j.energy.2021.121714

Under the complicated driving conditions, the sharp acceleration and deceleration actions would cause the high-rate charge and discharge current of electric driving system in hybrid electric vehicle (HEV), which brings about a serious impact on the battery lifetime. The hybrid energy storage system (HESS) combined with battery and ultracapacitor (UC), would be a possible solution to this problem. For HEV with HESS, in addition to improving fuel economy, realizing the protection of battery is also an important objective. However, improving one aspect performance may sacrifice another aspect performance. The tradeoff between multiple optimization objectives remains a challenge for energy management design. Aiming at this problem, a multi-objective optimization energy management strategy based on velocity prediction for a dual-mode power split HEV with HESS is proposed in this paper. Firstly, to get the precise predictive input sequence, generalized regression neural network (GRNN) is used to predict future velocity. Secondly, the power distribution of dual-mode power spilt HEV with HESS is described as a rolling optimization problem in the prediction horizon of model predictive control (MPC). A new cost function considering the fuel consumption and the protection of the battery is brought forward, and the optimization problem is solved using Pontryagin's minimum principle (PMP). Moreover, the Powell-Modified algorithm is introduced to execute the solving process of PMP. Finally, the proposed strategy is verified by comparing it with four other strategies under four different driving cycles. Compared to the rule-based strategy, the proposed strategy reduces root mean square (RMS) of battery current and fuel consumption by up to 18.5 % and 18.9 %, respectively. •Multi-objective optimization strategy is proposed for a dual-mode power split HEV.•The predictive control model is described with Pontryagin's Minimum Principle.•Power distribution over prediction horizon is solved using the Powell-Modified algorithm.•The performance is verified through the comparison among different methods.