Abstract
The accurate identification of permanent magnet synchronous motor (PMSM) parameters is the basis for high-performance drive control. The traditional PMSM multiparameter identification method experiences problems with the uncertainty of the identification results and low identification accuracy due to the under-ranking of the mathematical model of motor control. A multiparameter identification of PMSM based on a model reference adaptive system and simulated annealing particle swarm optimization (MRAS-SAPSO) is proposed here. The algorithm first identifies the electrical parameters of the PMSM (stator winding resistance R, cross-axis inductance L, and magnetic linkage ψf) by means of the model reference adaptive system method. Second, the result is used as the initial population in particle swarm optimization identification to further optimize and identify the electrical and mechanical parameters (moment of inertia J and damping coefficient B) in the motor control system. Additionally, in order to avoid problems such as premature convergence of the particle swarm in the optimization search process, the results of the adaptive simulated annealing algorithm to optimize multiparameter identification are introduced. The simulation experiment results show that the five identification parameters obtained by the MRAS-SAPSO algorithm are highly accurate and stable, and the errors between them and the real values are below 2%. This also verifies the effectiveness and reliability of this identification method.
Highlights
In recent years, permanent magnet synchronous motors (PMSMs) have been widely used in many area of production and life due to their many advantages, such as high power density, simple structure, and small size
We found that the model reference adaptive system algorithm (MRAS)-SAPSO algorithm be showed that the proposed multiparameter identification of the PMSM method based on the fied results closer to the true values among three algorithms, the five-p algorithm is feasible, and the quality the of the identification resultsand is higher
We proposed a step-by-step strategy to identify multiple parameters of a PMSM based on the MRAS-SAPSO algorithm, which effectively eliminates the problem of imprecise parameter identification caused by the coupling effect between the multiple parameters of a PMSM
Summary
Permanent magnet synchronous motors (PMSMs) have been widely used in many area of production and life due to their many advantages, such as high power density, simple structure, and small size. The neural-network-based convergence algorithm combined with the least square mean weight convergence has higher accuracy and faster convergence; the stability and speed of the recognition system depend on the selection of the convergence factor, and the neural network function approximation is sensitive to the training data [17,18] Such algorithms cannot estimate the rotor magnetic chain and winding resistance directly from the d–q equation of a conventional PMSM [19]. Zhang et al [25] and Feng et al [26] both proposed a method of injecting d-axis negative sequence current in a short time, which effectively solves the problem of under-ranking of the mathematical model of permanent magnet synchronous motors and can quickly achieve simultaneous multiparameter identification in. The simulated annealing algorithm was introduced to improve the optimal search strategy of particle swarm, which overcomes the limitation of particle swarm optimization falling into the local optimum, so produces a more accurate recognition effect
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
