DSP-based fault detection for DC-DC converters

Power electronics converter systems (PECS) are significant devices, which are used in industrial applications, including smart grids and variable speed AC drives. Therefore, the knowledge about the fault mode behavior of a converter system is extremely important from the perspective of protection and fault control in power systems. The present paper describes a novel on-line digital signal processor (DSP)-based diagnostic algorithm allowing the real-time detection, classification and localization of open-circuit (O-C) faults in the PECSs, as well as the identification of the unbalance input voltage to the converter. The proposed method requires much fewer input signals in comparison with the previous research works; therefore, the method avoids the use of additional sensors and signal processing devices that is important for the typical small-size commercial power converters used in distributed generation applications. Experimental results are presented using a coupled DC motor with AC synchronous generator scheme connected to the input of the power converter. The experimental results demonstrate the effectiveness and the robustness of the proposed diagnostic algorithms in this paper.