Naive Bayes Classifier for Cascaded H-Bridge Multi-Level Inverter Open Circuit Fault Diagnosis

This paper is explained open circuit fault detection and classification in Cascaded H-Bridge Multi-Level (CHBML) Inverter using designing Naive Bayes Classifier (NBC), thus enhancing efficacy and accuracy of the fault diagnosis. The pulse width modulation scheme is produces gate pulses in CHBML Inverter. The NBC is developed for detection and classification for open circuit fault employing the Haar Wavelets of voltages. Numerous open circuit faults have been analyzed in different CHBML switches, under parameter the effects. Moreover, the simulation results support the suggested NBC technique feasibility. Compared to the existing methods, the suggested NBC technique achieved a lower diagnosis time and given 98% open circuit fault classification accuracy. The suggested NBC technique is employed to any number of current and voltage levels. It can be noted that the NBC technique is more effective and can also be the benefit of being simple, reliable and fast based on different switch open circuit fault cases.