Search

Identifying and estimating uncertain and dynamic photovoltaic characteristics with high accuracy is important when modeling solar photovoltaic (PV) systems. It is critical to have effective and accurate parameters when transforming the complete PV system from solar energy to electrical energy. Therefore, we propose and apply for the first time a unique physics-based metaheuristic algorithm known as the Resistance–Capacitance Optimization Algorithm (RCOA), which is based on the concept of resistance–capacitance circuit response, and it was obviously inspired by the output response of a Resistor–Capacitor (RC) circuit when the input is connected or disconnected suddenly. The RC circuit's total time response can be divided into steady-state and transient-state. Both of these stages of the RC circuit are essential to model the algorithm properly. It is proposed that the RCOA be used as a single-objective algorithm that is simple, reliable, and has zero sensitive parameters, thereby being a parameter-free algorithm. The validity of the proposed RCOA is tested on the solar PV cell/module equivalent circuit parameter estimation of the three-diode PV model. The proposed algorithm combines an improved version of the Newton–Raphson method to find the optimal PV parameters in fewer iterations. The performance of the RCOA is compared with state-of-the-art algorithms, and the obtained results prove the superiority of the RCOA. The proposed methodology was found to be a trustworthy tool and it is proved through a statistical analysis and non-parametric test. Also, the results have revealed that the RCOA has superior reliability and accuracy when estimating the three-diode PV model parameters. It could be used as a viable approach for parameter identification problems in PV systems. With the average Friedman's ranking test value of 1.6666 and the average runtime of 17.08, the RCOA stands first among all selected algorithms. [ABSTRACT FROM AUTHOR]

When it comes to identifying the industrial solutions of solar photovoltaic (PV) systems, one of the difficult tasks is predicting the effectiveness of a PV system since the P–V, and I–V features of a PV system are nonlinear in nature. It is important to evidence that most manufacturers' specification sheets do not include comprehensive data about the equivalent circuit variables of PV models, which is essential to simulate an accurate solar module. Contrasted to other methods of extracting parameters from solar PV cells/modules, global research methodologies, and metaheuristic optimization algorithms are highly suitable as the main substitute for parameter extraction. Accordingly, this study proposes a new optimization approach for extracting the characteristics of solar PV cells/modules of various models, such as single-diode, double-diode, and module models, while accurately depicting the I–V and P–V curves. This is accomplished by using a chaotic generator in conjunction with the recently reported jumping spider optimization algorithm (JSOA) to obtain PV parameters from the optimization process. Comparative performance analysis reveals that the suggested optimization method, called CJSOA using a chaotic-based search sequence, performs better than many state-of-the-art optimization algorithms in terms of precision and reliability when identifying the PV parameters, as demonstrated in this paper. With the average Friedman's ranking test value of 1.397, the proposed CJSOA stood first among all selected algorithms. [ABSTRACT FROM AUTHOR]

Parameter identification and accurate photovoltaic (PV) modeling from basic I–V information are necessary for simulation, optimization, and control of the PV systems. Therefore, this paper proposes an Improved-African Vultures Optimization (I-AVO) algorithm, which combines the general Opposition-Based Learning (OBL) and Orthogonal Learning to extract the unknown parameters of the solar Photovoltaic (PV) modules accurately and effectually. The proposed I-AVO algorithm is developed from the basic version of the recently proposed African Vultures Optimization (AVO) algorithm. The solar PV parameters estimation problem is considered to be a complex optimization problem with the characteristics such as multi-dimensional, nonlinear, Transcendental, and multi-modal. Therefore, the basic variant of AVO struggles to produce the optimal and is stuck at local optima when it handles this complex optimization problem. Therefore, the I-AVO is formulated by combining the features of OL and OBL, along with the AVO, to generate the optimal solution. Out of various PV models, Three-Diode Model has been considered to determine the parameters. Furthermore, Newton–Raphson (NR) technique is discussed to solve the chaotic behavior of the I–V curve relation. The obtained results proved that the proposed I-AVO along with NR, called I-AVO-NR, can accurately obtain the optimal solution. The superiority of the proposed algorithm is proved to be better than other advanced algorithms based on the obtained results and their comparison. Based on the statistical test value obtained from Friedman's test, the proposed algorithm stood first among eight algorithms with the ranking value of 1.542 for two case studies. [ABSTRACT FROM AUTHOR]

In this study, an ABC-Local Search (ABC-Ls) method was proposed by including a new local search procedure into the standard artificial bee colony (ABC) algorithm to perform the parameter estimation of photovoltaic systems (PV). The aim of the proposed ABC-Ls method was to improve the exploration capability of the standard ABC with a new local search procedure in addition to the exploitation and exploration balance of the standard ABC algorithm. The proposed ABC-Ls method was first tested on 15 well-known benchmark functions in the literature. In the results of the Friedman Mean Rank test used for statistical analysis, ABC-Ls method successfully ranked first with a value of 1.300 in benchmark functions. After obtaining successful results on the benchmark tests, the proposed ABC-Ls method was applied to the single diode, double diode and Photowatt-PWP-201 PV modules of PV systems for parameter estimations. In addition, the proposed ABC-Ls method has been applied to the KC200GT PV module for parameter estimation under different temperature and irradiance conditions of the PV modules. The success of ABC-Ls method was compared with genetic algorithm (GA), particle swarm optimization (PSO) algorithm, ABC algorithm, tree seed algorithm (TSA), Jaya, Atom search optimization (ASO). The comparison results were presented in tables and graphics in detail. The RMSE values for the parameter estimation of single diode, double diode and Photowatt-PWP-201 PV module of the proposed ABC-LS method were found as 9.8602E−04, 9.8257E−04 and 2.4251E−03, respectively. In this context, the proposed ABC-LS method has been compared with the literature for parameter estimation of single diode, double diode and Photowatt-PWP-201 PV module and it has been found that it provides a parameter estimation similar or better than other studies. The proposed ABC-Ls method for parameter estimation of the KC200GT PV module under different conditions is shown in convergence graphs and box plots, where it achieves more successful, effective and stable results than GA, PSO, ABC, TSA, Jaya and ASO algorithms. [ABSTRACT FROM AUTHOR]