Scholarlink

<p>Power factor degradation and energy losses are significant challenges in electric vehicle (EV) charging stations. This study presents an IoT-enabled system with machine learning-based predictive analysis and dynamic reactive power correction to enhance power factor and reduce energy losses. The system utilized high-precision sensors, a Raspberry Pi-based edge computing unit, and nanostructured capacitors for reactive power correction. Real-time data was transmitted via the Message Queuing Telemetry Transport protocol to a cloud platform for monitoring and analysis. Machine learning models were trained on historical and simulated datasets to predict reactive power compensation dynamically. An optimization model was embedded within the control architecture to minimize reactive power deviation and ensure stable, near-unity power factor under varying load conditions. The system was evaluated under low (50&#xa0;kW), medium (100&#xa0;kW), and peak (150&#xa0;kW) load scenarios over 24-hour test periods. The system achieved substantial power factor improvement, from 0.85 to 0.99 in low-load, 0.83 to 0.99 in medium-load, and 0.80 to 0.98 in peak-load scenarios. Energy losses were reduced by 14.6%, 15.7%, and 18.4% under respective load conditions. Statistical analysis confirmed the effectiveness, with a paired t-test showing significant improvement (<i>p</i> &lt; 0.01) in power factor and ANOVA validating performance consistency across scenarios. Machine learning models demonstrated high predictive accuracy (R² = 0.97), with MAE of 0.03 and RMSE of 0.05. The system’s response times for load balancing were 15 ms, 18 ms, and 22 ms for low, medium, and peak loads, respectively. The IoT-enabled system efficiently optimized power factor and minimized energy losses in dynamic EV charging conditions. High predictive accuracy and rapid response times validate its practical viability. Future research should focus on scalability and integration with renewable energy sources for broader energy efficiency improvements.</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

IoT-enabled load power factor optimization in fast EV chargers

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年6月26日

查看全文
  • Priya,
  • Vimlesh Singh

摘要

Power factor degradation and energy losses are significant challenges in electric vehicle (EV) charging stations. This study presents an IoT-enabled system with machine learning-based predictive analysis and dynamic reactive power correction to enhance power factor and reduce energy losses. The system utilized high-precision sensors, a Raspberry Pi-based edge computing unit, and nanostructured capacitors for reactive power correction. Real-time data was transmitted via the Message Queuing Telemetry Transport protocol to a cloud platform for monitoring and analysis. Machine learning models were trained on historical and simulated datasets to predict reactive power compensation dynamically. An optimization model was embedded within the control architecture to minimize reactive power deviation and ensure stable, near-unity power factor under varying load conditions. The system was evaluated under low (50 kW), medium (100 kW), and peak (150 kW) load scenarios over 24-hour test periods. The system achieved substantial power factor improvement, from 0.85 to 0.99 in low-load, 0.83 to 0.99 in medium-load, and 0.80 to 0.98 in peak-load scenarios. Energy losses were reduced by 14.6%, 15.7%, and 18.4% under respective load conditions. Statistical analysis confirmed the effectiveness, with a paired t-test showing significant improvement (p < 0.01) in power factor and ANOVA validating performance consistency across scenarios. Machine learning models demonstrated high predictive accuracy (R² = 0.97), with MAE of 0.03 and RMSE of 0.05. The system’s response times for load balancing were 15 ms, 18 ms, and 22 ms for low, medium, and peak loads, respectively. The IoT-enabled system efficiently optimized power factor and minimized energy losses in dynamic EV charging conditions. High predictive accuracy and rapid response times validate its practical viability. Future research should focus on scalability and integration with renewable energy sources for broader energy efficiency improvements.