The change of motor parameters can affect the performance of permanent magnet synchronous motor (PMSM) position sensorless control. In order to solve this problem, the recursive least squares (RLS) algorithm is used for the online identification of PMSM parameters. Under the maximum torque per ampere (MTPA) control strategy, the improved model reference adaptive system (MRAS) based on back propagation (BP) neural network is used to construct a position sensorless control scheme, and a position sensorless control scheme for PMSMs based on parameter identification is proposed. The RLS method is used to identify the quadrature axis inductance and rotor flux of the PMSM online, and the parameter identification results are applied to the PMSM sensorless algorithm. Simulations and experiments prove that the parameter identification algorithm based on the RLS method can accurately identify the rotor flux and quadrature axis inductance of the PMSM, and the PMSM sensorless control scheme based on parameter identification is feasible.