Digital Twin Based Health Monitoring of Converter in Grid Integrated PV Systems using State Space Modeling and Advanced Control and Optimization Techniques

This research examines the application of an emerging technology known as the Digital Twin (DT) for the real-time health assessment of diverse grid-integrated photovoltaic (PV) system components. This study outlines the conception process, mathematical representation, and experimental validation of the newly constructed digital twin. The Digital Twin (DT) operates in conjunction with the physical system hardware, serving as an accurate mathematical model that offers insights for monitoring and regulation based on data obtained from the hardware's sensors. An innovative Rank-based Adaptive Artificial Electric Field Optimization (RAAEFO) algorithm is employed to match data gathered from the digital model with sensor readings from the actual system for data analysis. This paper introduces a novel Unit Vector Integrated Power Efficient Multi Constraint Model Predictive Control (UVIPEMC-MPC) methodology utilizing Adaptive Robust Cascaded Second-Order Generalized Integrator (AR-CSOGI) for regulating the transformation of energy within the specified system across multiple scenarios of operation. The computing environment utilized to execute the defined DT paradigm is OPAL-RT (OP4512). LABVIEW is utilized to execute the hardware prototype setup, operating on the NIsbRIO-9636 FPGA control board. Health depreciation evaluations of system components, including converter switches, interface inductors, and DC link capacitors, are conducted by assessing the effects of deterioration on various system characteristics. The efficacy of the created DT is assessed under multiple degradation levels. The recommended approach is executed on the hardware test product of the grid-integrated solar module.