Neural network based fault diagnosis and reconfiguration method for multilevel inverter

Multilevel inverter drivers have become widely applied in high-voltage and high-power applications. Therefore fault diagnosis of voltage source inverters is becoming more and more important. One possible fault within the inverter is IGBT open circuit fault. An overview of the different strategies to detect this fault is given, including the algorithms to localize the fault switch device. This paper presents a technique to improve the fault detection by using an algorithm of neural network with orthogonal basis functions based on recursion least square (RLS) technique. The method is used to identify the type and location of occurring faults from inverter output voltage measurement. Simulation examples of fault diagnosis by the use of the presented method were given. Simulation results have shown that the method performs satisfactorily to detect the fault type fault location and reconfiguration, so it will be very valuable in power system.