Hybrid Model-Based and Heuristic Optimal Parameter Estimation for a PV-Fed DC-DC Converter with Maximum Power Extraction

Precise estimation of photovoltaic (PV) parameters is crucial for extracting the Maximum Power Point (MPP). Heuristic MPP extraction methods suffer from the power oscillations phenomenon, whereas the model-based MPP results in convergence issues due to errors in estimated or measured parameters. In this article, a hybrid model-based and heuristic MPP extraction technique is proposed, where the heuristic approach for MPP extraction compensates for a reduced sensor estimation errors and oscillations in the model-based MPP extraction. The proposed methodology implements an iterative model-based PV parameter estimation to mitigate the above-mentioned drawback. The proposed algorithm estimates the PV equivalent circuit parameters at STC using the iterative technique with an initial guess. It is followed by estimation of irradiance and temperature and updating PV parameters. PV MPP voltage and current extraction is done using model-based iterative Guass-Seidel analysis and a dual-loop boost converter PI control loop. Higher-order polynomial approximation of the transcendental I-V equation is done for plotting PV curves using the estimated values and PV MPP voltage. Heuristric PV curves-based MPP extraction is aided with model-based MPP extraction to compensate for the errors. Simulation-based analysis is carried out for a PV module with a boost MPP converter, and the results for MPP extraction with sensorless estimation is provided for different loading conditions. Further, a prototype of 200 W boost converter is developed and validated for various irradiance and temperature profiles. Simulation and experiment results show the efficacy of the proposed technique for a Boost MPP converter.