A method for diagnosing multiple switch open failure of a three-phase PWM converter using artificial intelligence according to an embodiment of the present invention obtains a DC component (I V_dc ) of a dq-axis current in a stationary coordinate system, and a total harmonic distortion (Total Harmonic Distortion) of the current. , THD) (THD v ) estimating, applying the DC component (I V_dc ) of the dq-axis current to a pre-trained first artificial neural network (ANN) to determine the open failure type according to the current vector angle Grouping into a plurality of sectors and applying the DC component (I V_dc ) of the dq-axis current and the total harmonic distortion factor (THD v ) of the current to the second artificial neural network learned in advance to open failure within the grouped plurality of sectors and diagnosing the switch that has occurred. According to the present invention, the open failure of the switch is diagnosed using the angle of the current vector, but by using an artificial neural network (ANN) to newly classify the 21 open failure modes into 6 new open failure modes of multiple switches. Diagnosis is simple and accurate.
展开▼