Cauchy distribution optimizer with neighborhood search for PV parameter estimation and numerical optimization.
Abstract
This paper proposes an enhanced version of the Cauchy Distribution Optimizer (CDO) integrated with a Neighborhood Search (NS) mechanism for solving numerical optimization problems and extracting parameters of photovoltaic (PV) models. The original CDO, inspired by the Cauchy probability distribution, demonstrates competitive performance in balancing exploration and exploitation. However, like many metaheuristic algorithms, it may suffer from premature convergence and insufficient local search capability when handling complex multimodal problems. To address these limitations, we introduce a Neighborhood Search strategy that performs refined local exploration around promising solutions, thereby improving solution accuracy and convergence reliability. The proposed CDO-NS algorithm is validated using three PV models (Single Diode Model, Double Diode Model, and Photovoltaic Module Model). Experimental results demonstrate that CDO-NS outperforms the original CDO and other state-of-the-art algorithms in terms of solution quality, convergence speed, and statistical significance. The source code of CDO-NS(referred to as NSCDO throughout the paper) is made publicly available for further research.