Loss Minimization Control Based on Bivariate Extreme Value Theory for PMSMs

As considering both copper loss and iron loss, the loss minimization control (LMC) method could help permanent magnet synchronous motors (PMSMs) get higher efficiency compared with the traditional maximum torque per ampere control method. Unfortunately, it suffers from approximate models and the electromagnetic torque, which is hard to get precisely. To overcome these shortcomings and further improve efficiency, this article proposes an improved LMC method based on bivariate extreme value theory for PMSMs. First of all, based on the equivalent model of PMSM considering both copper loss and iron loss, the electrical loss of the PMSM is transformed into a function of the d-q -axis currents. Then, according to the bivariate extreme value theory, the optimal d-q -axis currents without the torque constraint are obtained. On this basis, the optimal d- axis active current model is derived, while the q -axis active current is obtained by the speed controller. After that a current regulator with the feedforward decoupling control of iron loss currents is designed to realize the LMC strategy with implicit torque constraint. At last, implementing comparative experiments with i d = 0 control and traditional LMC method, the improved LMC strategy can increase the efficiency of PMSMs significantly, which verifies its effectiveness.