The Study On Adjustable Speed System Of Brushless Doubly-Fed Machines Based On Fuzzy Neural Network

AC variable frequency speed regulation has a wide range of application in electric drive industry. In conventional induction motor adjustable speed drives, the cost of the power electronics substantially exceeds that of the machine alone and the resulting systems are expensive. Meanwhile, there exits a serious problem—harmonic current pollution, in the large bulk of the converter. The Brushless Doubly-fed Machines as a new type of electrical machine, costs much less than an induction motor system resulting from the lower rating converter, and the harmonic current is greatly reduced. BDFM can realize controllable speed operation precisely, which is simpler, its rotor structure being compact and robust that it can undertake arduous work. It has high valuable to put it into practice.In this paper, an analysis and study on BDFM is carried out. Firstly, the structure, operation principle, d-q mathematical model and parameter calculation method of BDFM are analyzed. Secondly, a simulation model in Matlab/Simulation software is built to verify dynamic, steady, and different operation mode's performance for BDFM.Fuzzy neural network combines the advantages of fuzzy logic system and neural network, which inherits not only the strong knowledge expression and logic reasoning abilities of fuzzy logic system, but also the powerful self-adaptive ability of neural network. In the paper, the application of fuzzy neural network(FNN) in Brushless Doubly-Fed Machines (BDFM) for adjustable speed drive are presented. This method applies the fuzzy neural network to the control of the closed loop of speed, tuning control rules real time with the change of the system. The simulation and experimental results show that the performance for the system are improved by using the FNN controller when the rotating speed or the load changes.