Comprehensive comparison of multi-objective optimization algorithms for extended-range electric vehicles incorporating noise, vibration, and harshness

• Published in: Journal of power sources Vol. 658; p. 238348
• Main Authors: Zhang, Yuxin, Yang, Yalian, Zou, Yunge
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.12.2025
• Subjects: Extended-range electric vehicles; Multi-objective energy management strategy; NVH control; Real-time control; Real-time control; Multi-objective energy management strategy; NVH control; Extended-range electric vehicles
• ISSN: 0378-7753
• DOI: 10.1016/j.jpowsour.2025.238348

Extended-range electric vehicles exhibit significant degradation in noise, vibration, and harshness (NVH) performance when operating in charge-sustaining mode. However, Current research primarily focuses on single-objective optimization, while the multi-objective co-optimization mechanism considering NVH characteristics and fuel efficiency remains underexplored. To bridge this gap, this study innovatively incorporates NVH characteristics into the optimization objective function, and based on the Pontryagin's Minimum Principle (PMP), Equivalent Fuel Consumption Minimization Strategy (ECMS), and Rule-Based strategies, a multi-objective collaborative optimization control algorithm incorporating the NVH performance is proposed - Fuel-NVH-PMP (FN-PMP), Fuel-NVH-ECMS (FN-ECMS), and Fuel-NVH-Rule-Based Strategy (FN-RB). Through a comprehensive comparative analysis of these three optimization algorithms, this study elucidates the inherent trade-offs between NVH suppression and fuel economy optimization. Furthermore, the adaptability of each control strategy is validated based on typical urban road conditions in Chongqing. Simulation results reveal that regarding NVH performance improvement, the FN-RB strategy requires a high fuel consumption cost and is not ideal overall. In contrast, the FN-PMP and FN-ECMS strategies achieve significant NVH performance improvements while maintaining fuel economy: NVH improves by 6–8 %, while fuel economy losses are only 1.3 % and 2.8 %, respectively. Finally, this paper constructs a hardware-in-the-loop experimental platform and validates the real-time performance of proposed strategy. •A multi-objective algorithm is designed to optimize fuel economy and NVH.•Real-time performance, NVH and fuel economy are comparatively examined.•Hardware-in-the-loop tests for online applications.