發(fā)布時間：2018-06-14 21:07

  本文選題：正交函數(shù) + 神經(jīng)網(wǎng)絡(luò)��； 參考：《控制與決策》2017年02期

【摘要】：針對含有多個未知非線性項(xiàng)的非線性系統(tǒng)難于控制的問題,提出利用Chebyshev正交函數(shù)構(gòu)建基于神經(jīng)網(wǎng)絡(luò)濾波器的控制器,并在權(quán)值學(xué)習(xí)誤差有界和跟蹤誤差有界條件下,通過李雅普諾夫穩(wěn)定性定理確定控制器的權(quán)值,保證了非線性系統(tǒng)的H∞魯棒控制.最后,利用所提出算法對非線性系統(tǒng)的濾波器和控制器進(jìn)行確定,仿真結(jié)果驗(yàn)證了該方法的有效性.
[Abstract]:To solve the problem that nonlinear systems with multiple unknown nonlinear terms are difficult to control, a controller based on neural network filter using Chebyshev orthogonal function is proposed, and the weighted learning error is bounded and the tracking error is bounded. The weights of the controller are determined by Lyapunov stability theorem, and the H 鈭，

本文編號：2018966