Research on single-pedal braking optimization based on hierarchical recognition of driving intention

Abstract

The single-pedal mode control has a significant impact on the energy efficiency and comfort of EVs. As a new type of integrated pedal system, changes in system characteristics and working modes pose higher requirements for optimizing human-machine interaction.Consequently, this paper conducts research on intelligent regenerative braking control based on the active recognition of single-pedal intentions. This paper proposes a hierarchical active recognition model for single-pedal intentions, RBF-GMM-HMM, which enhances the accuracy of driving intention recognition. Subsequently, based on the active recognition of driving intentions, it introduces optimal economic and coasting strategies for single pedals tailored to different operational modes and control domains. Finally, through full-vehicle and strategy modeling simulations, the paper verifies that the optimal economic strategy effectively improves the vehicle’s economy across various cycle conditions. Compared with the traditional strategy, the optimal economic strategy achieved 6.13%, 22.33% and 31.7% reduction of energy consumption under NEDC, UDDS, and US06 conditions; The optimal coasting strategy significantly enhances comfort while incurring relatively minor economic losses and ensures a constant coasting distance. Compared with the maximum coasting braking force strategy, the optimal coasting strategy only increases the SOC consumption by 0.75% under the premise of significant jerk optimization. The methods proposed in this paper better improve the applicability of single-pedal systems under different working conditions.