Font Size: a A A

Research On Fault Tolerant Control Based On BP Neural Network

Posted on:2006-09-15Degree:MasterType:Thesis
Country:ChinaCandidate:X P ZhaoFull Text:PDF
GTID:2168360155469650Subject:Control theory and control engineering
Abstract/Summary:PDF Full Text Request
Modern control system is being developed in the extensive and complicated scale. This kind of system is invested on a large scale. It may cause a great loss of personnel and property as soon as the accident takes place in this kind of system. Therefore it is important for people to improve the reliability and security of modern system. It is a new way that fault tolerant control is used to improve the reliability and security of modern system.Fault-tolerant control (FTC), which is developed in 1980's, is a kind of control technology in order to improve system's reliability. It ensures the system running stably and having perfect characteristics by using some effective methods, when one or more component of the system have failed or are about to fail. Fault-tolerant systems can generally be divided into two kinds: passive fault-tolerant control (passive FTC) and active fault-tolerant control (active FTC).ANN Fault Tolerant Control (ANN FTC) is the focus in the research of FTC recently. Modern control system is becoming more and more complicated, and it is difficult to gain the accurate model of modern system. And the system that is in fault is always nonlinear. It is a new way for people to use FTC based on neural network to resolve the problem of nonlinear and time-variable. The main contents of this paper are as follows:1. Modern system has many characteristics, such as non-linearity, large time lagging, time-varying and non-accurate mathematic model and so on, the author compares both advantages and disadvantages of different FTC methods, and the key point is put on the application of FTC basde on neural network.2. Using the improved BP algorithm, the controller and estimator of IMC based on neural network are designed.3. This paper brings FTC scheme directly using BP neural network as the fault compensator when the system is in trouble.4. Under the operation of fault compensator that the author has designed, this paper gives the sufficient condition of controller to guarantee the stability of closedloop system.5. If the system is in trouble, the system can achieve the aim of FTC when the fault compensator is added. The illustration at the end of the paper proves the validity of this designing method.
Keywords/Search Tags:BP neural network, BP algorithm, internal model control, fault- tolerant control
PDF Full Text Request
Related items