Font Size: a A A

Research On The Speed Sensorless Vector Control Of Induction Motor Based On Fuzzy Neural Network

Posted on:2012-05-15Degree:MasterType:Thesis
Country:ChinaCandidate:P G LiuFull Text:PDF
GTID:2212330338467235Subject:Rail transportation electrification and automation
Abstract/Summary:PDF Full Text Request
The speed sensorless vector control technology of induction motor is remarkable in reducing system complexity and cost effect. High-performance speed sensorless vector control system has become a hot research at home and abroad. In this paper, the classical speed identification method of model reference adaptive system is used to identify the speed. For nonlinear and variable parameters systems, the intelligent control methods- fuzzy control, neural network control and fuzzy neural network control were applied to improve the response and robustness of system effectively.In the thesis, depth studies the mathematical model of speed sensorless vector control system, the model of rotor flux observer, the principle of space vector pulse width modulation and fuzzy neural network theories. The paper focuses on how to use the fuzzy neural network regulator to replace the traditional PI speed regulator. Design and analysis the network structure, algorithms, and its stability. Based on the theoretical analysis, used the MATALB/SIMULINK tool to establish the simulation model and obtained many results. Simulation results show that algorithm can be used to meet the system performance; the system can get a good dynamic and static characteristics. In addition, this controller was compared with the traditional PI controller. The simulation results show that the fuzzy neural network speed controller can improve system performance obviously.For the parameter variations has a greater impact on system performance, we used fuzzy control and artificial neural networks to estimate the stator and rotor resistance respectively. Fuzzy control system doesn't need the accurate mathematical model of the control object and artificial neural network has superior ability to learn. The simulation results show that the parameter identification get a good effect.Finally, the impact of the iron loss on system performance was researched, considering the iron loss of motor model was analyzed, derived the iron loss compensation programs, established a control system considering iron loss. The simulation results show that with iron loss compensation of motor model can achieve the same output with the ideal motor model. It paves the way to realize the high-performance vector control system...
Keywords/Search Tags:speed sensorless, fuzzy neural network, MRAS, parameter identification, iron loss compensation
PDF Full Text Request
Related items