具有死区输入和时滞的未建模动态系统MNN控制  

Unmodeled dynamics and time delay systems with input dead zone MNN control

作　　者：李红春[1,2] 范丛山[1] 王斌[1] 

机构地区：[1]扬州工业职业技术学院电气信息工程学院,江苏扬州225127 [2]江海职业技术学院信息工程系,扬州225101

出　　处：《自动化与仪器仪表》2014年第10期115-118,共4页Automation & Instrumentation

基　　金：国家自然科学基金资助项目(61175111);江苏省高校"青蓝工程"(苏教师[2012]39号)资助

摘　　要：针对一类具有死区输入和时变时滞的未建模动态非仿射系统,基于滑模控制原理,提出了一种多层神经网络(MNN)控制器的设计方案。该方案通过构造适当Lyapunov-Krasovskii泛函有效地补偿了未知时滞不确定项。利用Nussbaum函数性质,放宽虚拟控制方向到未知。借助Young’s不等式,巧妙减少了传统多层神经网络中的一个自适应参数。理论分析证明了闭环系统半全局一致终结有界。仿真结果表明所提方案的有效性。Based on the principle of sliding mode control, an adaptive multi-input multi-output（MIMO） control scheme is proposed for a class of unmodeled dynamics non-affine systems with completely unknown control directions, unknown dynamics disturbances, unknown dead-zone and time-varying delay uncertainties. By constructing appropriate Lyapunov-Krasovskii functions, unknown time-varying delay uncertainties are effectively compensated for. And using the Nussbaum fimction properties, the unknown control directions are relaxed to be unknown. Further, Utilizing Young＇ s inequality, the parameters for traditional multi-neural network is reduced to two. By theoretical analysis, the closed-loop control system is proved to be semi-globally uniformly ultimately bounded. Simulation results demonstrate the effectiveness of the proposed method.

关 键 词：死区 神经网络控制 时变时滞 滑模控制 未建模动态 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]