Scholarlink

<p>Renewable energy systems, especially solar power, are vital for modern grids. As renewable energy sources increase, reactive power drawn from the grid rises, affecting power quality. This paper considers a single-phase grid-connected PV system with a new MPPT method, Spatial Recurrent PowerNet (SpaR-Net), for enhanced maximum power tracking and reactive power compensation. The SpaR-Net framework integrates spatial convolutional layers and a recurrent network for effective power tracking in unstable conditions. The reactive power is controlled using a hysteresis current controller, enabling smart inverter operation to deliver active and reactive power. Numerical results indicate that SpaR-Net performs better than other methods in accuracy (98%), power output (98.5%), and tracking efficiency (98.8%). SpaR-Net also provides a quick response time (0.2&#xa0;s). PSO-MPPT outperforms in power output and tracking efficiency, whereas P&amp;O and Fuzzy-MPPT have lower performances. SpaR-Net also exhibits the fastest convergence rate, i.e., improved adaptability towards irradiance changing conditions.</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

SpaR-Net Driven MPPT and Smart Inverter Control for Optimized Reactive Power Management in Solar PV Systems

  • 期刊论文

  • 发表日期：

    2026年5月4日

查看全文
  • Nirmala Rajendran,
  • Venkatesan Sundharajan,
  • Rajesh Kannan

摘要

Renewable energy systems, especially solar power, are vital for modern grids. As renewable energy sources increase, reactive power drawn from the grid rises, affecting power quality. This paper considers a single-phase grid-connected PV system with a new MPPT method, Spatial Recurrent PowerNet (SpaR-Net), for enhanced maximum power tracking and reactive power compensation. The SpaR-Net framework integrates spatial convolutional layers and a recurrent network for effective power tracking in unstable conditions. The reactive power is controlled using a hysteresis current controller, enabling smart inverter operation to deliver active and reactive power. Numerical results indicate that SpaR-Net performs better than other methods in accuracy (98%), power output (98.5%), and tracking efficiency (98.8%). SpaR-Net also provides a quick response time (0.2 s). PSO-MPPT outperforms in power output and tracking efficiency, whereas P&O and Fuzzy-MPPT have lower performances. SpaR-Net also exhibits the fastest convergence rate, i.e., improved adaptability towards irradiance changing conditions.