A novel online PMSM parameter identification method for electric and hybrid electric vehicles based on cluster technique

In order to save costs and volume, the torque is calculated based on a machine model instead of measuring it with a torque sensor. Thus, an accurate knowledge of the machine's electrical parameters is required for a high performance field-oriented control and torque accuracy. In this work, an innovative method is presented and developed for the online identification and correction of the electrical parameters of electrical machines based on cluster technique. This technique allows for the estimation of all electrical parameters with low computational cost. It uses the information of stator currents, stator voltages, and rotor angular speed. The algorithmic performance has been tested using simulation data of a 75-kW permanent magnet synchronous motor for electric vehicles. The results confirm the high accuracy of the identification method. This approach is applicable to electric and hybrid electric vehicles.