Core Loss Effect Modeling and Compensation for Improved MTPA Control of PMSM Drive under High-Speed Conditions

For permanent magnet synchronous machines (PMSMs), accurate machine model is critical for high performance maximum torque per ampere (MTPA) control. However, as motor speed increases, the nonlinearity such as core loss effect will affect the accuracy of machine model and thus the performance of online MTPA control. This paper firstly investigates the performance of the model based MTPA control under different motor speeds through modeling, simulation and experiments, which indicates that the accuracy of MTPA control is greatly reduced especially under high-speeds due to machine nonlinearity. Hence, this paper proposes an efficient nonlinearity compensation model based on polynomial fitting to model and compensate the MTPA error as motor speed increases. Considering both core loss and magnetic saturation effects, the compensation model is a nonlinear polynomial of speed and stator current. To obtain the fitting data, a derivative modeling method is proposed to compute the actual and detected MTPA angles under different speeds, in which the derivative model of torque to current ratio is fitted and the MTPA angle is obtained by setting the derivative model to zero. The proposed compensation model is both computation effective and easy to use for MTPA control, as it computes the compensation term that can be directly combined to other model-based methods. The proposed model is evaluated with experiments and comparisons on a test motor to show the performance improvement.