Research And Application Of The Direct Torque Control On Brushless Doubly-fed Machine

Brushless Doubly-Fed Machine (BDFM) is a new kind of machine, which can operate in asynchronous, synchronous or doubly-fed models. It can be used in AC adjustable-speed drive systems as a motor or be used in generating systems as a generator. The BDFM system has the characteristics as follows: less capacity inverter, separately adjustable active power and reactive power, adjustable power factor, continuously adjustable speed in four quadrants, brushless operation and so on, so it will have broad applications in saving-energy-adjusting-speed systems of wind machines and pump and variable-speed-constant-frequency wind power and water power generating systems.According to the difference of rotor structure, Brushless Doubly-Fed Machine mainly has two types: Brushless Doubly-fed Cage Machine and Brushless Doubly-fed Reluctance Machine, in this dissertation the main research-object is BDFRM. In the view of the air gap field of BDFM, the operating principles and basic structures of stator and rotor are introduced in detail, at the same time a d-q-0 model has been developed under rotating reference frame. With the dynamic equation, three kinds of operating model, synchronous,asynchronous and doubly-fed, are investigated. The analysis of simulation results indicate that the machine possess excellent dynamic performance under these operating models. By use of MATLAB/Simulink, a conventional direct torque control model is constructed, then the paper deeply analyze the advantage and disadvantage of this control strategy, simulation results and theory analysis keep high accord, therefore it can be said that the simulation model of BDFM is right.Aiming at the shortcoming of the conventional direct torque control, the thesis proposed a novel method to improve dynamic characteristic of machines by application of intelligent control to replace bang-bang control. Firstly, a converter switch-table is established by means of fuzzy control, then the parameters of fuzzy membership functions and rule-bases table have been optimized by through of genetic algorithm. Genetic Algorithm is a kind of excellent intelligent algorithm, it possess great advantage on aspect of function optimizing in comparison with other algorithm, yet it also has some unsatisfactory aspects, such as: premature,low local search efficiency etc. Many researchers do some work to improve the capability of genetic algorithm, in this thesis Chaos Algorithm and Immune Algorithm have been used to meliorate genetic algorithm. The adjustable thickness trait of immune algorithm can avoid premature phenomenon; chaos algorithm's good characteristics, randomicity,ergodicity,regularity, can improve local search efficiency, in this way genetic algorithm will quickly reach its best point. At last, a intelligent control model of BDFM has been built by application of MatLab/Simulink, simulation results validate the Chaos Immune Genetic Algorithm and machine's capacity has a great improvement.At last, the dissertation introduced the application of brushless doubly-fed machines in wind power generation. It can be believed that the research accomplished by the paper has a important practical value for the application of direct torque control strategy in the area of wind power generation.