基于神经网络PID算法的研究与设计

传统PID的算法模型对于复杂的过程和非线性系统难于建立准确的数据模型,甚至在控制过程中出现不稳定现象.为解决此类问题,按PID控制方式选择神经元学习控制所需的状态变量,将自适应神经元与ID控制算法结合起来,较好地解决了交流变频调速系统负载扰动、时滞现象及非线性控制的问题.通过仿真实验比较传统PID和NN PID算法,说明NN PID具有较好的鲁棒性.

Study and Design on PID Algorithm Based on Neural Network

As for intricate process and non-linear system,the traditional PID algorithm mode has trouble in setting up accurate data mode and caused non-steady phenomenon.The solution is to establish PID control algorithm of adaptive neural network,offering adaptive learning algorithm that solves problems such as load disturbance of AC frequent control system,time lag phenomenon and non-liner control.Finally,the paper compared PID and NN PID algorithm by an emulating experiment which proved good robustness of NN PID.