Scholarlink

<p>The fast global adoption of renewable energy systems requires smart and flexible control strategies to make solar photovoltaic (PV) installations as efficient as possible in changing environmental conditions. This study proposes a hybrid Adaptive Neuro-Fuzzy Inference System–Random Forest (ANFIS–RF) based Maximum Power Point Tracking (MPPT) technique for high-performance PV systems. The proposed controller combines the nonlinear approximation capability of ANFIS with the ensemble learning robustness and fast decision-making of the Random Forest algorithm through an adaptive decision-level fusion mechanism. PV voltage, current, irradiance, and temperature are employed as input features, while the estimated maximum power point is used to regulate the PV operating voltage. Comprehensive simulations are conducted on a 540 kWp rooftop PV system subjected to rapidly varying irradiance conditions ranging from 600 to 1000&#xa0;W/m². The proposed method is benchmarked against conventional RF, ANN, and standalone ANFIS MPPT techniques. Results demonstrate that the hybrid ANFIS–RF controller achieves a peak MPPT efficiency of 99.8%, outperforming artificial neural network (ANN) (98.9%), ANFIS (98.5%), and RF (97.3%). Power ripples are reduced to 1.2%, with the fastest settling time of 1.0&#xa0;s and the lowest RMSE of 190&#xa0;W. The instantaneous tracking ratio exceeds 0.95 for 89% of operation, confirming strong robustness and stability for smart grid PV applications.</p>

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

A next-generation AI-driven random forest MPPT control architecture for solar energy extraction

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月17日

查看全文
  • J. Viswanatha Rao,
  • Jagarapu S. V. Siva Kumar,
  • Juvvanapudi Sharon Rose Victor,
  • Kiran Kumar Pulamolu,
  • Deepthi Janyavula,
  • Rajanand Patnaik Narasipuram

摘要

The fast global adoption of renewable energy systems requires smart and flexible control strategies to make solar photovoltaic (PV) installations as efficient as possible in changing environmental conditions. This study proposes a hybrid Adaptive Neuro-Fuzzy Inference System–Random Forest (ANFIS–RF) based Maximum Power Point Tracking (MPPT) technique for high-performance PV systems. The proposed controller combines the nonlinear approximation capability of ANFIS with the ensemble learning robustness and fast decision-making of the Random Forest algorithm through an adaptive decision-level fusion mechanism. PV voltage, current, irradiance, and temperature are employed as input features, while the estimated maximum power point is used to regulate the PV operating voltage. Comprehensive simulations are conducted on a 540 kWp rooftop PV system subjected to rapidly varying irradiance conditions ranging from 600 to 1000 W/m². The proposed method is benchmarked against conventional RF, ANN, and standalone ANFIS MPPT techniques. Results demonstrate that the hybrid ANFIS–RF controller achieves a peak MPPT efficiency of 99.8%, outperforming artificial neural network (ANN) (98.9%), ANFIS (98.5%), and RF (97.3%). Power ripples are reduced to 1.2%, with the fastest settling time of 1.0 s and the lowest RMSE of 190 W. The instantaneous tracking ratio exceeds 0.95 for 89% of operation, confirming strong robustness and stability for smart grid PV applications.