Longitudinal–vertical comprehensive control for four-wheel drive pure electric vehicle considering energy recovery and ride comfort

Highlights: • A longitudinal–vertical interaction model of electric vehicle braking is created. • A comprehensive control method combining model predictive and neuro-fuzzy control is proposed. • A neuro-fuzzy optimization framework is proposed. • Double-loop multi-stage control is selected for comparison. • Comprehensive control improves energy recovery efficiency and ride comfort. The energy recovery efficiency and ride comfort of electric vehicles are important performance indicators. Currently, few joint studies have been conducted on the energy recovery and ride comfort of electric vehicles. For enhanced energy recovery and ride comfort, a comprehensive control method that contains neuro-fuzzy control and model predictive control is proposed herein. First, a longitudinal–vertical interaction model under braking conditions is established that includes longitudinal–vertical variables interaction. Second, model predictive control is adopted to adjust the active suspension for improving ride comfort with the braking intensity as the disturbance. Subsequently, to improve the energy recovery efficiency of the vehicle, a neuro-fuzzy optimization framework is proposed for optimizing the neuro-fuzzy membership function to realize neuro-fuzzy control, the framework considers the constraints of the vehicle vertical motion on the braking torque. Furthermore, the neuro-fuzzy control is adopted to control the vehicle powertrain. Finally, a dual-loop multi-stage control is selected for comparison. The simulation results under combined braking conditions indicate that the proposed comprehensive control method simultaneously improves the energy recovery efficiency and ride comfort of the vehicle.