基于BP神经网络的PID算法研究与应用

摘要

针对实际非线性、不确定系统，文章将提出基于BP算法神经网络PID算法，阐述了三层BP网络设计方法。最后用于一个慢变性被控对象进行仿真，结果表明，此控制方法能够有效克服经典 PID 控制器在被控对象具有非线性、时变不确定性和难以建立精确的数学模型时出现的参数整定不良等问题。%According to the actual nonlinear, uncertain control system, the paper will introduce BP neural network PID algorithm and Described the three layers BP network design method .at last use this control method on the Slow variant controlled object, simulation result showed that This control method can effectively overcome the problem such as nonaccuracy of classic PID controller of non-linear, time-varying uncertain control system.