Parameter estimation of PMSM driven by PWM inverter based on discrete dynamic model

Abstract

Most PMSM parameter estimation techniques assume steady state to avoid model error caused by current derivative but have difficulty in collecting data rich enough to estimate four parameters of PMSM in regular operation. Parameter estimation based on dynamic model has advantages such as simultaneous estimation of four parameters, data collection from regular operation and faster estimation. In this research, PMSM parameter estimation based on the discrete dynamic model using data synchronously sampled at the top and bottom of triangular carrier of a PWM inverter is proposed. The influence of model errors on the estimation due to discretization method, dead time and resolution of PWM are investigated in a simulation study. Immunity of the proposed estimation to the dynamic change in current is tested in the experiment