Battery independent regenerative braking using model predictive control with auxiliary power consumers

Abstract

Electric vehicles (EV) are able to regenerate the energy of approximately 85% of all braking procedures. In rare cases, e.g. if the battery is fully charged, the EV is not able to regenerate. Therefore, the regenerative braking system is not part of the service brake. Due to its heavy battery, the EV brakes are even larger than those of a conventional vehicle. This paper analyzes the potential of downsizing the friction brake by using auxiliary power consumers like the high-voltage heater (HVH), the air-conditioning system (AC) and low-voltage auxiliaries, taking into account their dynamic behavior. The cabin temperature has to be nearly constant for driver comfort. To ensure the assumptions made regarding the auxiliary consumers, measurements were taken from an EV (VW e‑Golf). Further, measurements of an air- and water-heating system were conducted. A model predictive control (MPC) strategy was used to distribute the power in an optimal way in consideration of the system dynamics. The MPC uses quadratic programming to optimize a cost function containing the dissipated brake energy and the cabin temperature deviation for a linearized model. The overall upshot of this feasibility study is the general capability of battery-independent recuperative braking in EVs and its effect on friction brake downsizing. The importance of the systems transient response was evaluated with a MPC control strategy.