运用常规的PID控制算法很难达到人们所要求的控制效果。采用改进的BP神经网络算法进行改进具有以任意精度逼近非线性函数的能力，而且通过它的自身的学习，可以找到某一最优控制率下的PID控制器参数，使其具有更好的鲁棒性和自适应的能力。-Using conventional PID control algorithm is difficult to live up to the required control effect. The improved BP neural network algorithm to improve accuracy with an arbitrary nonlinear function approximation ability, but also through its own learning, you can find an optimal control rate of PID controller parameters, to make it better robustness and adaptive capacity.