The cost functions of model predictive torque control (MPTC) of permanent magnet synchronous motor (PMSM) need to be weighted and summed because of the different control objective dimensions, but these are difficult to design and adjust the weight coefficients. The objective evaluation methods such as fuzzy decision method, VIKOR method, TOPSIS method, coefficient of variation weighting method and entropy weighting method are applied to the model predictive torque control of permanent magnet synchronous motor to solve the above problems. Examining the model predictive torque control of permanent magnet synchronous motor with flux linkage and torque control, the above methods are feasible and do not need weight coefficients, but the calculation load is increased to a certain extent. There are some differences in the control performance of different strategies, but the overall control performance is basically the same. According to the comprehensive performance comparison, the control performance of TOPSIS method is relatively optimal.