Scholarlink

<p>The cumulative implementation of Electric Vehicles (EVs) has driven the need for efficient and sustainable power management solutions. Therefore, this work proposes an integrated Photovoltaic (PV)-based EV charging system featuring a high-gain quadratic boost-Single Ended Primary Inductor converter (SEPIC) to enhance power conversion efficiency. To maximize the PV power extraction under both uniform and partial shading conditions, a Kepler-optimized Recurrent Neural Network (RNN) is employed for Maximum Power Point Tracking (MPPT). The optimized RNN ensures faster convergence and higher accuracy in tracking the global peak power, minimizing oscillations and power loss with the incorporation of optimization topology. The system utilizes a bidirectional DC-DC converter to regulate energy flow towards the EV battery and DC bus, ensuring efficient charging and discharging processes. Proportional-Integral (PI) controllers regulate the current and voltage at various stages, enhancing system stability and protecting the EV battery from overvoltage and overcurrent conditions. This system validates the effectiveness by implementing in MATLAB/Simulink, which signifies higher efficiency of (97.5%), tracking accuracy (99.95%) and minimal oscillations based on the comparative analysis made with the proposed topology. This outcomes a reliable and scalable solution for EV charging infrastructures with higher performance.</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

Efficient PV-Integrated EV Charging Station with Quadratic Boost-Sepic Converter and Optimized RNN for Enhanced Voltage Conversion

  • 期刊论文

  • 发表日期：

    2025年9月16日

查看全文
  • I. Mahendravarman,
  • C. Srisailam,
  • Krishna Prakash Arunachalam,
  • D. Lakshmi

摘要

The cumulative implementation of Electric Vehicles (EVs) has driven the need for efficient and sustainable power management solutions. Therefore, this work proposes an integrated Photovoltaic (PV)-based EV charging system featuring a high-gain quadratic boost-Single Ended Primary Inductor converter (SEPIC) to enhance power conversion efficiency. To maximize the PV power extraction under both uniform and partial shading conditions, a Kepler-optimized Recurrent Neural Network (RNN) is employed for Maximum Power Point Tracking (MPPT). The optimized RNN ensures faster convergence and higher accuracy in tracking the global peak power, minimizing oscillations and power loss with the incorporation of optimization topology. The system utilizes a bidirectional DC-DC converter to regulate energy flow towards the EV battery and DC bus, ensuring efficient charging and discharging processes. Proportional-Integral (PI) controllers regulate the current and voltage at various stages, enhancing system stability and protecting the EV battery from overvoltage and overcurrent conditions. This system validates the effectiveness by implementing in MATLAB/Simulink, which signifies higher efficiency of (97.5%), tracking accuracy (99.95%) and minimal oscillations based on the comparative analysis made with the proposed topology. This outcomes a reliable and scalable solution for EV charging infrastructures with higher performance.