基于线性二次型的单神经元PID最优控制器设计及仿真

THE DESIGN AND SIMULATION OF THE SINGLE-NEURON PID OPTIMAL CONTROLLER BASED ON LINEAR QUADRATIC

讨论了线性二次型最优控制理论在神经网络PID控制系统中的应用。将二次型性能指标引入单神经元PID控制器中,设计出单神经元自适应PID最优控制器,从而实现了PID参数在线自适应寻优。给出了完整的设计过程和学习算法,分析了其稳定性。最后,运用MATLAB仿真实现证明了该方法的可行性、有效性,并得到了较为理想的控制效果。

This paper discusses the application of linear quadratic optimal control theory in network PID control system. The quadratic index was introduced into single-neuron PID controller and then the optimal controller was designed for accomplishing PID parameters＇ online adaptive optimization. The whole algorithm and design process were outlined, the stability was analyzed too. Finally the simulation results on MATLAB proved the feasibility and effectiveness of this method and the ideal control effect was achieved.