The engine friction and actuator power are the main factors of the cooling system affecting the fuel economy of a spark ignition (SI) engine. An electrified cooling system containing an electric fan, pump, and thermostat provides an opportunity to reduce fuel consumption. The coolant temperature is always kept at a high fixed value within the safe temperature range to avoid friction losses caused by overcooling; however, the actuator power is not typically considered. Recent publications have attempted to minimize the actuator power and the coolant temperature is maintained in a range. Nevertheless, neither method quantitatively considers both factors. In this paper, the integrated consideration of engine friction and actuator power is presented to minimize engine fuel consumption. The accuracy of a control-oriented model of a cooling system is improved first in an attempt to exert the full potential of the model. Then, the proposed strategy for minimum fuel consumption is constructed as an optimization problem and the improvement of fuel economy obtained by the proposed strategy is evaluated using a causal suboptimal controller and dynamic programming (DP)-based global optimal controller. Compared with a causal coolant temperature tracking controller, the causal suboptimal controller and the global optimal controller based on the proposed strategy both achieve significant improvements. Compared with a global optimal controller for minimum actuator power, the global optimal controller based on the proposed strategy achieves a certain improvement and this effect can increase as the environmental temperature decreases. Finally, a real-time implementation of the proposed strategy on a hardware platform using model predictive control (MPC) with a limited horizon is presented, which shows the feasibility of the proposed strategy.
Read full abstract