基于模型预测的纯电动汽车动力总成热管理策略

Thermal management strategy of battery electric vehicle powertrain based on model predictive control

对某款纯电动汽车的动力总成热管理系统进行研究,基于AMEsim软件搭建分布式冷却系统,包括驱动电机、电机控制器冷却系统和动力电池冷却系统,对模型中的参数进行了匹配,基于实际台架对模型进行了验证,证明了搭建模型的准确性。基于热管理系统的状态空间模型,在Simulink软件中设计模型预测控制器,以热源的发热功率作为扰动量,冷却水泵流量和风扇转速作为操作变量对冷却系统的控制策略进行了优化,通过和AMEsim软件的联合仿真,结果表明相比PID控制和开关阈值控制,采用模型预测控制的热管理系统响应性能更好,对于温度的变化会提前作出反应,同时也减少了冷却系统的能量消耗。

Based on AMEsim software distributed cooling system was built,including the drive motor,motor controller and power battery cooling system,the powertrain thermal management system of a certain type of pure electric vehicle was studied.The parameters in the model were matched,the model was proved to be accurate based on the actual test bench.Based on state space model of thermal management system,the model predictive controller was designed in the Simulink software.Taking the heating power of heat source as disturbance,cooling water pump flow and fan speed as operating variables,the control strategy of the cooling system was optimized through joint simulation of Simulink and AMEsim.The results showed that the thermal management system with model predictive control had better response performance,compared with the PID control and switching threshold control.It also reduced the energy consumption of the cooling system.