Improved Parameter Estimation of Three-Phase Squirrel-Cage Induction Motors Using the Nelder-Mead Simplex Algorithm

Abstract

This work presents a technique for precisely determining the characteristics of a squirrel-cage three-phase induction motor using the Nelder-Mead simplex algorithm. This approach is a frequently employed numerical optimization technique for determining the minimal value of a multi-dimensional objective function. An advantageous feature of the Nelder-Mead simplex algorithm is its independence from the need to calculate partial derivatives of the objective function. Nevertheless, similar to several optimization techniques, the Nelder-Mead simplex approach can also exhibit sensitivity to the initial conditions. Thus, the initial estimation of the parameters of the approximated equivalent circuit of the induction motor was used as the starting point for the Nelder-Mead optimization approach. The experiment's results are compared to those obtained using the polynomial regression approach to demonstrate the efficacy of the proposed method.