Scholarlink

<p>Maximum power moment tracking (MPPT) methods sometimes miss the ideal photovoltaic (PV) system optimization moment due to rapid sun irradiance and temperature changes. A hierarchical optimization framework for a Zeta converter–based photovoltaic system uses a modified perturb-and-observe technique, Fuzzy TOPSIS, an adaptive neuro-fuzzy inference system (ANFIS), and long short-term memory (LSTM) networks to solve this task. The coordinated decision–learning–prediction mechanism uses multicriteria fuzzy decision analysis to rank operational points, adaptive nonlinear inference to refine them, and short-term environmental forecasting to alter them. Fuzzy TOPSIS identifies operating positions around the ideal maximum-power area using real-time irradiance, temperature, voltage, and current. ANFIS models the nonlinear interactions between environmental inputs and converters to maximize duty-cycle correction. The long short-term memory network predicts near-future trends in irradiance and temperature, enabling converter adjustment before power output is compromised. The layered technique transforms maximum power point tracking (MPPT) from reactive tracking to predictive control at the global maximum power point under dynamic conditions. Tracking accuracy, response time, and energy output are significantly higher than those of typical controllers in steady and rapidly changing environments. Maximum power point tracking efficiency averages over 96% and peaks over 98% under changing conditions. The system’s rapid convergence following irradiance transients, reduced steady-state oscillations, and enhanced resilience during partial shading boost daily energy production. An integrated intelligent solar optimization control system balances robust decision-making, nonlinear flexibility, and predictive capability. The technique increases the reliability, efficiency, and sustainability of grid-connected and freestanding solar energy systems by enabling continuous operation across wide voltage ranges and under environmental disturbances.</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

Sustainable Model for Optimization of Power Operation of Photovoltaic System with Zeta Converter Based on Hybrid Machine Learning Techniques

  • 期刊论文

  • 发表日期：

    2026年3月24日

查看全文
  • Rupali S. Balpande,
  • Rashmi Keote,
  • Ujwala S. Ghodeswar,
  • Vaishali P. Raut,
  • Lowlesh N. Yadav,
  • Tejas R. Patil,
  • Nischal Puri,
  • Rohit Pawar,
  • Yoginee S. Pethe,
  • Vikrant S. Vairagade

摘要

Maximum power moment tracking (MPPT) methods sometimes miss the ideal photovoltaic (PV) system optimization moment due to rapid sun irradiance and temperature changes. A hierarchical optimization framework for a Zeta converter–based photovoltaic system uses a modified perturb-and-observe technique, Fuzzy TOPSIS, an adaptive neuro-fuzzy inference system (ANFIS), and long short-term memory (LSTM) networks to solve this task. The coordinated decision–learning–prediction mechanism uses multicriteria fuzzy decision analysis to rank operational points, adaptive nonlinear inference to refine them, and short-term environmental forecasting to alter them. Fuzzy TOPSIS identifies operating positions around the ideal maximum-power area using real-time irradiance, temperature, voltage, and current. ANFIS models the nonlinear interactions between environmental inputs and converters to maximize duty-cycle correction. The long short-term memory network predicts near-future trends in irradiance and temperature, enabling converter adjustment before power output is compromised. The layered technique transforms maximum power point tracking (MPPT) from reactive tracking to predictive control at the global maximum power point under dynamic conditions. Tracking accuracy, response time, and energy output are significantly higher than those of typical controllers in steady and rapidly changing environments. Maximum power point tracking efficiency averages over 96% and peaks over 98% under changing conditions. The system’s rapid convergence following irradiance transients, reduced steady-state oscillations, and enhanced resilience during partial shading boost daily energy production. An integrated intelligent solar optimization control system balances robust decision-making, nonlinear flexibility, and predictive capability. The technique increases the reliability, efficiency, and sustainability of grid-connected and freestanding solar energy systems by enabling continuous operation across wide voltage ranges and under environmental disturbances.