Hardware Design and Implementation of a Fault-Detection in DC/DC Converters Based on Kalman Filtering

With the current changes in the automotive sector, especially in electric vehicles, various architectures have emerged at different voltage levels and with this diversity, the importance and number of uses of DC/DC power converters are increasing. The stability and reliability of these converters, which have become more critical, and the gradual improvement of system health management are changing the perspective of solutions and applications in automotive architectures. This thesis focuses on the degradation of the output filter capacitor of power converters and is based on an algorithm that will provide real-time monitoring and anomaly detection and provide accurate data about the stability of power converters to the user or system provider. In full terms, it includes hardware design and application development on the algorithm. The printed circuit board (PCB), which was created following the feasibility and efficiency studies verified by simulations, was the starting point for the production of the first prototype. This application board was planned to allow the DC/DC power converter to operate at different frequencies, optionally at High Frequency and Low Frequency. In this case, additional capacitors based on the parallel architecture in the output capacitors and activated provide flexibility to the circuit. A system is created that starts with the separation of the AC component in the output signal from DC and continues with the Analog filter. The data is first digitized by means of the Analog Digital Converter (ADC) and then transferred to the Field Programmable Gate Array (FPGA). The codes tested via Vivado are integrated into the FPGA and drive together with the PCB. This hardware offers the opportunity to interpret test results from different perspectives with its compact structure and architecture that provides sufficient flexibility for prototypes.