Parameter Identification of Asynchronous Motors Based on Improved Recursive Least Squares and Model Reference Adaptive Method

The traditional recursive least squares (RLS) method involves first and second-order derivatives of voltage and current signals in the parameter identification model for asynchronous motor parameters. However, the results of direct discretization are significantly influenced by noise, leading to relatively poor accuracy in rotor resistance identification. To address this issue, an online parameter identification method for induction motors based on an improved recursive least squares (RLS) algorithm and model reference adaptive system (MRAS) is proposed in this paper. This method effectively reduces the influence of truncation error by employing a Butterworth digital filter and solving the intermediate variables of the state equation using the fourth-order Runge-Kutta method. Additionally, it corrects the identified rotor resistance by integrating the model-reference adaptive system. Simulation and experimental results demonstrate that, compared to existing methods, the proposed approach achieves higher accuracy, with no more than a 1% error in identifying each motor parameter.