Font Size: a A A

The Application Of Adaptive Inverse Controlin Industrial Process

Posted on:2008-06-06Degree:MasterType:Thesis
Country:ChinaCandidate:Y PangFull Text:PDF
GTID:2178360215980714Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
Adaptive inverse control is a new kind of control technology and is the result of combination between adaptive control and adaptive signal processing, which was put forward by Widrow professor of Stanford University in 1986. The advantage of this control technology is that it can separate the design of controller and disturbance canceller. The application of adaptive inverse control in linear system is rather mature, but the application in nonlinear just seeds. So this paper focuses on its application in nonlinear system. Widrow brought forward a kind of control structure of adaptive inverse control, which has been applied in some linear control systems, but the necessary condition of canceling disturbance in this structure is the linear addition of controlled system. So this structure could not apply well in nonlinear system. Aiming at this problem, three kinds of nonlinear control system structure are put forward in this paper. They are introduced respectively as followed:Nonlinear adaptive inverse control based on inverse system is the first structure. Inverse system is firstly introduced into adaptive inverse control area in this paper. Moreover, support vector machine is used to model theαorder inverse system of controlled system, which is located before the controlled system and the controlled nonlinear system is transformed to false linear system, which can be controlled with linear system adaptive inverse control theory. What is more, another advantage of inverse system is that it can disconnect the relation among variables and economy the number of filter. The results of simulations show the perfect the performance of this structure.Nonlinear adaptive inverse control based on hybrid neural network is the second. The combination of nonlinear static network and linear dynamic network is used to model the nonlinear system. When modeling finished, the structure of nonlinear static and linear dynamic system has been determined, then work out the inverse system of nonlinear static and linear dynamic system with support vector machine. According as deduced formation, an nonlinear adaptive inverse control system is composed. The results of simulations show the perfect the performance of this structure.Nonlinear control system based on separable model is the third. The idea of this structure is decomposing a system into an nonlinear static model and linear dynamic model and figuring out the inverse model respectively. But this structure differs with the second structure essentially. Because this structure controls the nonlinear system through controlling the linear dynamic part of nonlinear system instead of transforming the nonlinear system into linear system, so this system is linear control system essentially. The results of simulations show the perfect the performance of this structure.
Keywords/Search Tags:Adaptive inverse control, Inverse system, LS-SVM, Hybrid neural network, nonlinear adaptive control, Separable model
PDF Full Text Request
Related items