Convergence analysis and corresponding strategy of full model based induction motor flux observation

In high performance AC system, whether the flux position can be obtained correctly or not will influence the dynamic and steady state performance. Induction motor full model realizes motor speed on-line identification as the parameter, besides motor current and flux real-time observation as state variables simultaneously. Research is done on the convergence characteristics of flux observation in the process of estimating motor parameters and variables. Due to the nonlinearities of full model, singularity theory is adopted, dividing full model into speed identification model and flux observation model, with the two models separated on time scale. By analyzing the eigenvalue distribution and damp ratio of flux observation subsystem, convergence and influence factors are studied. Research results show that the damp ratio in the middle and high speed range needs to be improved and the corresponding strategy of improving convergence is proposed. Simulation and experiments verify the validity of analysis method and the feasibility of proposed strategy.