Abstract: The electric vehicle has more time-varying, nonlinear and large time lag characteristic, and traditional proportion integration differentiation (PID) control method cannot achieve the desired control effect. For the back propagation (BP) neural network, deducing the formula of the forward direction and reverse transmission, giving the detailed design of control algorithm, designing a kind of BP neural network PID controller on the basis of the traditional PID control. Through the simulation of heat pump air conditioners model, and results show that the controller enables the system with high stability, good robustness etc, and is superior to the traditional PID control effect. Finally, the BP-PID algorithm and pulse width modulation (PWM) control are combined and the hardware and software principle of the system are designed, compared with the traditional PID control, the stability time is reduced from 155 s to 145 s to realize the temperature control of heap pump air conditioners, preparing for future vehicle development.

Key words: Electric vehicle; BP-PID; PWM; Heat pump air conditioners; Temperature control

摘要： 电动汽车热泵空调系统具有时变性、非线性和滞后性的特点，传统的比例-积分-微分 （PID）控制方法无法达到理想的控制效果。针对反向传播（BP）神经网络，推导出其正向 和反向传播阶段公式，给出了详细的控制算法设计，在传统 PID 控制器基础上设计出一种自 学习 BP 神经网络 PID 控制器。对热泵空调系统模型仿真，结果表明，该控制器具有稳定性 高，鲁棒性好等优点，优于传统 PID 控制效果。最后把 BP-PID 算法与脉冲宽度调制（PWM） 控制相结合，进行了系统软硬件原理设计，与传统 PID 控制相比稳定时间从 155 s 减少到 145 s， 实现空调温度控制，为后续车型开发做准备。

关键词: 电动汽车；BP-PID；PWM；热泵空调；温度控制