A Novel KSK Converter with Machine Learning MPPT for PV Applications

Abstract

Power electronic converters are regarded as indispensable aspects of modern electric power systems, owing their eminence as optimal power interface and power flow regulators. Hence, the design of an efficient converter is vital to support wide-scale expansion in the application of both Renewable Energy Sources and Electric Vehicles. Thereby, a novel high gain K. S. Kavin (KSK) converter based on interleaved architecture, exhibiting lower voltage stress across switches is proposed in this research work. The suggested converter design is well suited for Photovoltaics (PV) power generation, as it offers higher step-up gain with lower input current ripples. Moreover, mathematical analysis and modeling of suggested KSK converter with different operational modes are also covered in this article. Additionally, the Radial Basis Function Neural Network based on Machine Learning is employed as a Maximum Power Point Tracking technique to optimize power generated by the PV system. Additionally, Internet of Things (IoT) is used for real-time monitoring of PV parameters. The viability of the suggested KSK converter is demonstrated through MATLAB simulation and laboratory prototype implementation, while the control of the developed 1000 W prototype is entrusted up on the FPGA Spartan 6E controller. Consequently, achieved exceptional 98.6% efficiency of the proposed KSK converter lends credence to its excellence over other published converter topologies.