Research On Speed Sensorless Control Of Six Phase Induction Motor Based On Adaptive Observer

With the fast development of full-controlled semiconductor device, microprocessor and digital signal processor, three-phase induction machine drive system has been greatly developed. Compared to traditional three-phase induction machine, multiphase machine is characterized as:be able to realize higher power output under low voltage, smaller torque ripple and higher stability. Thus multiphase machine has prospected in applications of high power drive system where supplied voltage amplitude is limited. Even so, the application areas of multiphase induction machine are still limited today, because theoretical researches on multiphase induction machine just have been start and not mature enough. Thus it has gradually become a hot research focus to study multiphase induction. At present, the amount of researches on speed estimation of multiphase machine are still small. Under these circumstances, some researches on sensorless vector control of six-phase induction machine basing on research of speed estimation method for three-phase machine has been conducted in this thesis.The main research object of this thesis is six-phase induction machine. Basing on whose windings space feature, the construction of mathematical model has been analyzed in detail, which includes the voltage, current, flux and kinematic equations, also the decouple coordinate transformation matrix which reflects the transformation from six-phase natural system of coordinates to two-phase static coordinates. After that, the mathematical model of six-phase induction machine built in two-phase static coordinates has been introduced. In this thesis, the UVM vector control of three-phase induction machine also has been analyzed, and combined with features of six-phase induction machine, the UVM vector control of six-phase induction machine finally has been built.The characters of model reference adaptive control and sliding-mode observer has been analyzed in this thesis. Basing on methods of speed estimation of three-phase induction machine, adjustable models of these two kinds of speed estimation methods have been constructed. By adjusting parameters of PI, it can be realized that the rotor flux error between adjustable model and reference model down to zero. Model reference adaptive control and sliding-mode observer both have strong anti-jamming capability, and well static and dynamic performance. Besides, for model reference adaptive control method, it is capable in a wide range of speed estimation, and it has shown strong robustness to parameters variation of induction machine.With the help of simulation software MATLAB/Simulink, the vector control system model of six-phase induction machine, which basing on model reference adaptive control and sliding-mode observer methods, have been built. In the simulation model, the performance of vector control system and speed estimation results under various kinds of working situation have been simulated, thus to verified the theory proposed above.At the end of this thesis introduced the experiment part. Digital signal processor has been used as center control core. Main power circuit and hardware control circuit of induction machine vector control system have been built. And software control algorithm of six-phase induction machine vector control system has been designed. Finally, experiment has been conducted to verified the performance of six-phase induction machine vector control system basing on model reference adaptive method, The experiment also helped in making preparation for later research work.