三级倒立摆的GA-PIDNN系统辨识

GA-PIDNN System Identification for Triple Inverted Pendulum

针对典型的不稳定、多变量、非线性、强耦合的三级倒立摆系统,建立了基于GA优化的PID神经网络(GA-PIDNN)辨识结构,完成了GA与BP两种算法的简单对比,并给出了MATLAB仿真结果.结果表明,GA-PIDNN对于非线性三级倒立摆的辨识是有效的,且GA优于BP算法.

Aiming at the typical instability, multi-variable, nonlinear, strong-coupling of triple inverted pendulum system, the PID neural network based on GA optimization （GA-PIDNN） identification structure is established. The simple comparison of GA and BP algorithm is completed. MATLAB simulation results are presented. The results show that GA-PIDNN identification of nonlinear triple inverted pendulum is effective, and the GA is better than BP algorithm.