说明：  基于BP神经网络的PID控制器结构,控制器由两部分组成：一是常规PID控制器，用以直接对对象进行闭环控制，且3个参数在线整定；二是神经网络NN，根据系统的运行状态，学习调整权系数，从而调整PID参数，达到某种性能指标的最优化。
(基于BP神经网络的PID控制器结构,控制器由两部分组成：一是常规PID控制器，用以直接对PID controller structure BP neural network, the controller consists of two parts: First, a conventional PID controller for closed loop control of the direct object, and three-line parameter tuning the second is the neural network NN, operate in accordance with the system state, learning to adjust the weights to adjust the PID parameters to achieve certain performance optimization.)