A driving-style-oriented adaptive control strategy based PSO-fuzzy expert algorithm for a plug-in hybrid electric vehicle

• Published in: Expert systems with applications Vol. 201; p. 117236
• Main Authors: Lin, Xinyou, Li, Kuiliang, Wang, Liming
• Format: Journal Article
• Language: English
• Published: New York Elsevier Ltd 01.09.2022; Elsevier BV
• Subjects: Adaptive control; Algorithms; Consumption; Driving style recognition; Electric vehicles; Energy management strategy; Equivalence; Expert systems; Fuel consumption; Fuel economy; Fuzzy control; Fuzzy expert algorithm; Hybrid electric vehicles; Neural networks; Particle swarm optimization; Plug-in hybrid electric vehicle; Principal components analysis; Fuzzy expert algorithm; Driving style recognition; Energy management strategy; Particle swarm optimization; Plug-in hybrid electric vehicle
• ISSN: 0957-4174; 1873-6793
• DOI: 10.1016/j.eswa.2022.117236

•A driving styles recognition algorithm based two layers Fuzzy controller is defined.•An equivalent factor PI-fuzzy updating rules is applied for adapting driving styles.•A PSO-Fuzzy algorithm is incorporated into the driving style-oriented strategy.•The proposed strategy can improve the fuel economy for various diving styles. The propelling torque of a plug-in hybrid electric vehicle (PHEV) is provided by the engine and driving motor. However, the driver style has a great influence on the fuel economy performance of the PHEV. To address this issue, this study proposes a novel driving-style-oriented adaptive control strategy by using Particle Swarm Optimization (PSO) combined with Fuzzy expert algorithm. The first is development of a driving style recognition method by taking advantage of the Fuzzy expert algorithm. The four input parameters of Fuzzy expert system are derived by Principal Component Analysis based the collected data of driving behavior real tests. Second, a novel driving-style-oriented adaptive control strategy is developed based the Equivalent Consumption Minimization Strategy (ECMS) integrated with adaptive Equivalent Factor (EF). The EF is updating according to the corresponding driving style by using the PI-Fuzzy controller, and the PSO algorithm is applied to further optimize the fuzzy rules for tuning the PI factors. Finally, in combination with the above efforts, a novel driving-style-oriented adaptive control strategy based PSO-Fuzzy Expert Algorithm has been further established accordingly. A normal ECMS with constant EF and an adaptive ECMS for neural network to recognize the driving style is applied to compare with the proposed strategy. The validation results indicate that the strategy significantly decreasing the fuel consumption with the equivalent fuel consumption is reduced by 19.3%, and the proposed strategy has the best performance.