Parameter Identification of Permanent Magnet Synchronous Machine Based on an Adaptive Mutation Dynamic Differential Evolution

The problem of parameter estimation of permanent-magnet synchronous machines (PMSMs) can be formulated as a nonlinear optimization problem. To obtain accurate machine parameters, it is necessary to develop easily applicable but efficient optimization algorithms to solve the parameter estimation models. This paper proposes a novel dynamic differential evolution with adaptive mutation operator (AMDDE) algorithm for the multi-parameter simultaneous estimation of a non-salient pole PMSM. The dynamic updating of population enables AMDDE to responds to any improved changes of the population immediately and thus generates better optimization solutions compared with the static mechanism used in original differential evolution. Two mutation strategies, namely DE/rand/1 and DE/best/1, are adaptively employed to balance the global exploration and local exploitation. The effectiveness of the proposed AMDDE is demonstrated on the multi-parameter estimation for a non-salient pole PMSM. Experimental results indicate that the proposed method significantly outperforms the existing peer algorithms in efficiency, accuracy and robustness. Furthermore, the new algorithm can be potentially realized in real-time microcontroller due to its simple structure and less memory requirement. The proposed algorithm can also be applied to other parameter identification and optimization problems.