Research On Off-line Identification Of Induction Machine Parameter

High performance induction machine speed control system has been widely used in industrial and agricultural production. As the key technologies for control system, flux linkage estimation and parameter identification are becoming attractively. The study in this paper has been done under this circumstance.In this paper, different parameter identification schemes have been summarized. After the analysis and comparison of various strategies, the thesis combined the advantages of MRAS(Model Reference Adaptive System) and EKF(Extend Kalman Filter) and chose this two methods finally. The model of an induction motor with MRAS parameter identification and EKF rotor resistance identification has been done based on Matlab/Simulink respectively. Four parts of the research work are briefly listed as follows:Firstly, according to the state space equation of induction motor, the induction motor model has been created using S-function in MATLAB, and the simulation results show that this model has a good dynamic and steady performance.Secondly, this paper does research on parameter identification using MRAS which was based on Lypunov theory. The result gradually converge to it's true value by selecting parameters'adaptive law and dynamic adjustment. If the excited condition has been satisfied accurately, stator resistance R s, rotor resistance R r, stator inductance L s and rotor inductance can be identified accurately. And leakage inductance ratioδand rotor time constant Tr can be obtained using formula calculation synchronously. The paper has studied above method under stator and rotor reference axis respectively, which included deducing the adaptive laws of parameters and proving the convergence of parameters theoretically. In order to confirm the operation performance of the estimation model, a whole simulation in Maltab/Simulink has been done in this paper. The simulation results show that this method was correct and effective.Thirdly, because the rotor resistance is one of the most obvious changeable parameter for the induction motor control system. In order to improve the robustness of the control system, the rotor resistance must be identified rapidly and accurately. In this paper, a novel algorithm of estimating the rotor resistance of induction motor(IM) based on EKF has been studied. The method utilizing double closed-loop controlling strategy can do resistance identification on line by measuring stator voltage, currents and rotor speed. A simulation model system is built by using Matlab/Simulink, and the shift of the rotor resistance influence on the speed has been analyzed. The results show that the validity and accuracy of the presented algorithm.Finally, stator voltage, stator current and rotational speed under stable operation has been collected using oscillograph. The collected voltage and current data has been deal with spectrum analysis and filtering processing, then load in simulation model for calculation, the result show that the identification model has a good convergence.