基于Elman神经网络的二自由度内模控制方法研究

RESEARCH ON TWO-DEGREE-OF-FREEDOM INTERNAL MODEL CONTROL BASED ON ELMAN NEURAL NETWORK

将遗传算法、Elman神经网络和二自由度内模控制相结合,提出一种基于Elman神经网络辨识和遗传整定的二自由度内模控制算法。仿真结果表明,相对于常规内模控制方法,该方法对输入信号的动态响应快,系统超调量和调节时间减少,抗干扰能力强,控制精度高。

With the combination of genetic algorithm,Elman neural network and two-degree-of-freedom internal model control,a two-degree-of-freedom internal model control algorithm based on Elman neural network recognition and genetic tuning is proposed.Simulation results show that,compared with traditional internal model control,the proposed algorithm improves the response speed against input signals,reduces the system overshoot and settling time,and has strong ability in interference rejection and high control precision.