Scholarlink

The precise modeling of the photovoltaic (PV) modules is necessary for performance forecasting, diagnostics, and the design of control systems. In this paper, we propose a Monte Carlo-based optimization method for estimating the parameters of a single-diode PV model using experimental current-voltage (I-V) data. The method takes advantage of stochastic sampling to explore large parameter spaces and determine the best parameters that provide the least error on assumed versus measured data. Among other parameters, the optimized values yielded the following results: photocurrent Iph was 0.77483 A, reverse saturation current I0 was estimated as 6.57 × 10−7 A, ideality factor n of 1.606, series resistance Rs of 0.084 Ω while shunt resistance Rsh was 74.64 Ω. The resulting model yielded remarkable agreement with experimental data, achieving a maximum simulated power output of 0.2920 W at 0.414 V. Model accuracy was confirmed by a high coefficient of determination (R2 = 0.9722) and low error values (RMSE = 0.0503 A, MAE = 0.0376 A). Residual analysis validated the lack of systematic bias while parameter sampled histograms displayed stable convergence behavior. Researchers using traditional heuristic optimization strategies will notice that Monte Carlo methods are more straightforward and considerably more robust; these methods are advantageous for noisy datasets. The results confirm the applicability of the method.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Optimization of Single-Diode Photovoltaic Model Parameters Using Stochastic Optimization and Real Experimental Data

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Oussama Khouili,
  • Fatima Wardi,
  • Mohamed Louzazni,
  • Mohamed Hanine

摘要

The precise modeling of the photovoltaic (PV) modules is necessary for performance forecasting, diagnostics, and the design of control systems. In this paper, we propose a Monte Carlo-based optimization method for estimating the parameters of a single-diode PV model using experimental current-voltage (I-V) data. The method takes advantage of stochastic sampling to explore large parameter spaces and determine the best parameters that provide the least error on assumed versus measured data. Among other parameters, the optimized values yielded the following results: photocurrent Iph was 0.77483 A, reverse saturation current I0 was estimated as 6.57 × 10−7 A, ideality factor n of 1.606, series resistance Rs of 0.084 Ω while shunt resistance Rsh was 74.64 Ω. The resulting model yielded remarkable agreement with experimental data, achieving a maximum simulated power output of 0.2920 W at 0.414 V. Model accuracy was confirmed by a high coefficient of determination (R2 = 0.9722) and low error values (RMSE = 0.0503 A, MAE = 0.0376 A). Residual analysis validated the lack of systematic bias while parameter sampled histograms displayed stable convergence behavior. Researchers using traditional heuristic optimization strategies will notice that Monte Carlo methods are more straightforward and considerably more robust; these methods are advantageous for noisy datasets. The results confirm the applicability of the method.