Self-learning-pattern-based method and system for preventing hand clamping during door closing of motor

The invention provides a self-learning-pattern-based method and system for preventing hand clamping during door closing of a motor. The system comprises a microprogrammed control unit (MCU), a control circuit, a motor, a sampling circuit and an analog/digital (A/D) conversion circuit, wherein the sampling circuit acquires a loop voltage value of a loop of the motor; and the MCU receives the loop voltage value by the aid of the A/D conversion circuit and controls the motor by the aid of the control circuit. According to the method, in the door opening process, the maximum loop voltage value of the motor is acquired to be used as a locked-rotor valve value, in the door closing process, the loop voltage value of the motor is acquired and is compared with the locked-rotor valve value, whether a locked-rotor condition occurs is judged, if the locked-rotor condition occurs, the motor stops door closing, and if the locked-rotor condition does not occur, the motor closes a door continuously. With the method and the system, the problem of inconsistent locked-rotor valve values due to differences of motor consistency, performance change after long-time usage and the like can be effectively solved, the locked-rotor detection sensitivity is greatly improved, and personal safety is really guaranteed.