Aiming at the problem of model parameter mismatch leading to control system performance degradation in permanent magnet synchronous motor under complex working conditions, a method of differential beatless predictive current control (DPCC) based on online identification of parameters is proposed. Firstly, the predictive current control model of PMSM is established, and the sensitivity of each motor parameter to the conventional DPCC is analyzed in detail. Secondly, a motor parameter recognizer is designed by using the method of Adaline neural network; on this basis, a variable step-size neural network weight adjustment algorithm applied to the motor system is proposed; finally, the online recognized parameters are used to update the parameters in the current prediction controller in real time, in order to avoid the impact of parameter mismatch on the performance of the control system. It is proved through experiments that the proposed method can not only achieve accurate online tracking of motor parameter changes, but also effectively improve the convergence speed of the identification results.