Rule-Based Alternator Control Using Predicted Velocity for Energy Management Strategy

The market concern of improvement of vehicle safety and its convenience to drive a vehicle has resulted in the growth of the demand for vehicular electronic equipment. This trend requires additional power in the vehicle and thus makes prone to the increase of fuel consumption for vehicles equipped with internal combustion engines. To minimize this fuel consumption, an efficient energy management (EM) strategy for the electrical system of alternator and battery is required. This paper proposes a successful EM strategy based on the rule-based alternator control using predictive information. The proposed strategy reduces fuel consumption by charging batteries using the residual kinetic energy during deceleration. In particular, we predict electrical energy that is recovered by the residual energy using a Markov chain-based velocity prediction algorithm. The accommodation of predicted electrical energy and current vehicle information determines one of the three predefined control modes, such as charge, hold, and discharge, depending on vehicle driving states. This control mode determines the power generation from the alternator and controls the amount of torque to the vehicle electrical system. The proposed strategy is verified through simulation and experiment. The simulation with the new EM strategy is validated as comparing the operation difference with a conventional proportional-integral (PI) control algorithm under the same driver behaviors. Further validation in real vehicle driving experiment shows that fuel consumption was reduced by 2.1% compared to the conventional PI control algorithm.