A new control algorithm of regenerative braking management for energy efficiency and safety enhancement of electric vehicles

A new generation of electric vehicles with in-wheel motor technology has been introduced and developed. Increasing system efficiency, eliminating mechanical intermediaries, and achieving regenerative braking torque with better performance are the motivations to seek to improve this technology. To the best knowledge of authors, no solution has been introduced to enhance such system efficiency while all constraints and limitations, especially the battery health and safety issues have been considered. In order to properly manage the energy and use the maximum possible braking capacity of the motor's regenerative mode, accurate instantaneous information of the maximum possible braking torque capacity is required. In the present study, a half-car model with five degrees of freedom has been developed by considering a vehicle equipped with two in-wheel motors on the rear axle as a sample vehicle. The braking strategy has been designed using a two-stage nonlinear predictive controller. The first stage will reach the appropriate pressure for the brake fluid lines. Then, in the second stage, the proper amount of electric regenerative torque is obtained using the electronic braking force distribution (EBD) function and considering all constraints. The amount of regenerative torque is calculated by considering the system constraints using the Karush-Kuhn-Tucker conditions (KKT). Finally, the designed strategy is examined from the perspective of safety and vehicle mileage capability. The results show that optimal braking can be achieved, and the braking stopping distance can be reduced by utilizing the designed controller and the proposed model. Also, the amount of regenerated energy to the battery can be increased during braking by using the proposed braking strategy and the designed control system compared with the relevant studies. The proposed strategy enhanced the vehicle mileage by more than 24 percent while considering all constraints and limitations.