Font Size: a A A

Fault Diagnosis And Fault Tolerant Control Of Switched Reluctance Motor

Posted on:2007-11-09Degree:MasterType:Thesis
Country:ChinaCandidate:W H GanFull Text:PDF
GTID:2192360182478945Subject:Power electronics and electric drive
Abstract/Summary:PDF Full Text Request
Switched Reluctance Drive is a kind of new electrical drive that has the special double-salient pole structure characteristics. It has excellence of both asynchronous motor and direct current motor drive systems, that is the wide speed range and the excellent performance. This virtue makes its extensive application in home electric appliances, universal industry, aviation industry, servo and driving system, traction motor and high speed motor systems etc.In order to improve the reliability and stability of the system, the thesis research on the fault diagnosis of the drive system, and takes the one phase short and one phase open fault as research object.The thesis introduced the basic structure characteristics of the switched reluctance drive system, the necessity and the present research status of fault diagnosis. Analyzed the development of the switched reluctance drive fault diagnosis, and introduced the popular intelligent diagnosis technology, such as Neural Network. Then analyzed the often fault of the SRD and gene rant mechanism of the faultStudied and established the mathematic model of the SRD. Modeling is the basic of the other research works such as the control strategy and the fault diagnosis. Introduced the present research status and the problems of modeling SRD, then established the linear model of the SRD. After deeply analyzed the inner electromagnetism relation of the SRD, established the nonlinear model, and put forward the simulating and analysis result.Introduced the neural network technology which is fairly popular in fault diagnosis, also emphases the BP network, which is the fault diagnosis method in this thesis. Based on the sample data of the fault, constructed the three layer BP network structure, and applied the standard back-propagation, variable learning rate back-propagation and conjugate gradient arithmetic to train the network, then used the testing sample data to test and diagnose the network. The testing result shows that the three layer BP network has preferable diagnosis capability, and the conjugate gradient arithmetic is the feasible arithmetic.At last, the thesis studied primarily about the winding short circuit fault of .the SRD.
Keywords/Search Tags:Switched Reluctance Drive, Fault Diagnosis, Nonlinear Model, Neural Network, Back-Propagation Arithmetic, Open Phase Fault, Short Phase Fault
PDF Full Text Request
Related items