SISO仿射非线性系统自适应变结构BPNN控制  

Variable Structure BPNN Adaptive Control for SISO Affine Nonlinear Systems

作　　者：邹劲松[1] 唐旭[2] 

机构地区：[1]重庆水利电力职业技术学院电子信息工程系,重庆402160 [2]重庆大学科技处,重庆400044

出　　处：《计算机测量与控制》2015年第2期457-459,共3页Computer Measurement &Control

摘　　要：为了更好地消除抖振,提高复杂非线性系统的控制效果,针对一类典型SISO仿射非线性系统,提出了一种新的变结构BP神经网络(back propagation neural network,BPNN)自适应控制策略(VSYNC);其中,对于系统未知非线性函数,将神经网络用作估计器;对控制输入加入连续函数项,其可以根据状态点和滑动切换面之间的距离自适应地调节不连续的控制变量,从而使变结构控制策略得到了显著提高;所提出的控制方法能够有效地抑制周围切换面的抖振,保证了系统地动态性能,同时还能够消除系统的稳态误差;仿真结果表明,该方法具有较高的控制精度和鲁棒性。In order to better eliminate the chattering and improve the control effect of complex nonlinear system,a novel adaptive control strategy of variable structure neural network （VSYNC） for a kind of typical SISO affine nonlinear systems is proposed.Among them,the neural network is used as the estimator for unknown nonlinear function systems.This strategy can add continuous function items for the control input,and can adaptively adjust discontinuous control variables according to the distance between the state point and the sliding switching surface,so that the variable structure control strategy will be improved significantly.The proposed control method can effectively suppress the chattering of switching surfaces around,ensure the dynamic performance of the system,and can also eliminate steady-state errors of the system.Simulation results show that the method has high control accuracy and robustness.

关 键 词：非线性系统 变结构控制 神经网络 自适应 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]