基于神经网络广义逆的两电机变频系统内模控制  

Internal Model Control of Two Motor Variable Frequency Systems Based on Neural Network Generalized Inverse

作　　者：刘国海[1] 杨官学[1] 沈跃[1] 陈兆岭[1] 

机构地区：[1]江苏大学电气信息工程学院,镇江212013

出　　处：《电工技术学报》2010年第11期56-61,67,共7页Transactions of China Electrotechnical Society

基　　金：国家自然科学基金(60874014);教育部博士点基金(20050299009);江苏省自然科学基金(BK2007094)资助项目

摘　　要：针对非线性强耦合的两电机变频系统,先利用动态BP神经网络逼近原系统的广义逆系统,然后再对复合后的伪线性系统提出了基于广义逆系统的内模控制方法,证明了闭环系统的鲁棒稳定性。最后基于S7?300PLC的平台,对系统设计做了相关的试验研究。结果表明,神经网络广义逆系统方法,不但能够很好地实现系统的解耦,而且还可以使伪线性化后的子系统开环稳定,引入的内模控制,又保证了系统的控制性能。For nonlinear high coupling two motor variable frequency systems（TMVFS）,the dynamical BP neural network is firstly used to approximate to generalized inverse system of original system,then an internal model control（IMC） based on generalized inverse system method is proposed to this pseudo-linear system.Stability and robustness of the closed-loop system are proved,and its performances are analyzed.Finally,some relative experimental researches based on S7-300 PLC are studied.The results demonstrate that decoupling control with open-loop stability can be reached by neural network generalized inverse（NNGI） system method and the control performance of system can be ensured by inducting internal model control to subsystems.

关 键 词：两电机系统 BP神经网络 广义逆 内模控制 解耦控制 鲁棒稳定性 

分 类 号：TM351[电气工程—电机]