Sensorless control can reduce the cost of permanent magnet motor and improve the reliability of the system. Because the inductance and other parameters of permanent magnet motor change with the operating conditions, the accurate identification of motor parameters has a positive significance to improve the control accuracy of motor. Based on the mathematical model of permanent magnet motor, the parameters of motor inductance are identified by artificial neural network, and then the rotor angle is observed by sliding mode observer (SMO) to realize the sensorless vector control of permanent magnet motor. Using TI′s TMS320F28379d DSP as the control chip, a motor control circuit is built to verify the above control strategy. The results show that the control strategy can accurately identify the motor parameters in real time, as well as improve the control precision and performance.