Study And Implementation On A Sensorless Vector Control Technology Of Induction Motors Based On Flux Observer

This paper proposes a new method for sensorless vector control of induction motors, which is based on the flux observer theory and the adaptive control theory. According to the vector control theory of an induction motor, we utilize the speed estimation and speed control by the rule that the rotor flux of the voltage model must be coincident with the one of the current models. In order to solve the problems of flux estimation using the voltage model, such as the initial value and the drift of pure integrator, a flux-observer-based method is applied. The advantages of this system are that the fluxes from the current model are easily obtained, all of the fluxes (from voltage and current models) are dc quantities, which are convenient for control and checking, and the system stability is easily analyzed. A further advantage is the avoidance of integration problems. In order to improve the dynamic performance of this system, this paper proposes an improved speed estimator, which is based on a variable argument PI mechanism. The linear model of the proposed system is derived to analyze the system stability and to select the controller gains.Additionally, this paper proposes an experimental system using DSP (TMS320F240) as a main controller to realize the system mainframe design and some of the functional programs. Meanwhile, this paper also includes some of the peripheral circuit design. Furthermore, this paper introduces the dynamic simulation design of speed sensorless vector control system using MATLAB/SIMULINK.Finally, the paper presents the stability analysis results and dynamic simulation results of the proposed system to prove the system validity. Meantime, we also offer the variable argument PI control simulation results contrast to the ordinary PI control's results to verify its effectiveness.