Scholarlink

<p>Congestion analysis is critical for power system analysis and operation, but integration of energy storage systems (ESSs) and renewable energy sources (RESs) brings great uncertainty and computational challenges. This study presents an advanced approach that integrates a Mix-Style Neural Network (MSNN) for high-accuracy load forecasting with the Starfish Optimization (SFO) algorithm for optimal ESS scheduling under varying grid conditions. The major objective is to enhance system reliability, reduce operational and congestion-related costs, and improve adaptability during N-1 contingency events. Unlike traditional methods, the proposed strategy creates direct feedback loop between load forecasting and ESS control, enabling more responsive and cost-effective decisions. MSNN captures nonlinear demand variations, while SFO ensures efficient convergence for complex scheduling. Validation using the IEEE 24-bus system was carried out in MATLAB platform, demonstrating that proposed method outperforms established techniques as Chaos Game Optimization (CGO), Improved Particle Swarm Optimization (IPSO), and Improved Manta Ray Foraging Optimization (IMRFO), achieving 45.73% reduction in congestion management costs, prediction accuracy of 98.5%, and computation time of 3&#xa0;seconds. Furthermore, it delivers the most favorable economic performance with a levelized cost of energy (LCOE) of 0.13$/kWh and net present cost (NPC) of 185,627$. Statistical results confirm significant improvements in prediction accuracy and efficiency. The proposed framework contributes to smarter, more resilient grid management under renewable uncertainty.</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

Adaptive congestion mitigation in smart grids using MSNN-SFO framework for energy storage system scheduling under renewable uncertainty

  • 期刊论文

  • 发表日期：

    2025年10月7日

查看全文
  • K. Paul Joshua,
  • Beedalannagari Omprakash,
  • Mukuloth Srinivasnaik,
  • Chinthalacheruvu Venkata Krishna Reddy

摘要

Congestion analysis is critical for power system analysis and operation, but integration of energy storage systems (ESSs) and renewable energy sources (RESs) brings great uncertainty and computational challenges. This study presents an advanced approach that integrates a Mix-Style Neural Network (MSNN) for high-accuracy load forecasting with the Starfish Optimization (SFO) algorithm for optimal ESS scheduling under varying grid conditions. The major objective is to enhance system reliability, reduce operational and congestion-related costs, and improve adaptability during N-1 contingency events. Unlike traditional methods, the proposed strategy creates direct feedback loop between load forecasting and ESS control, enabling more responsive and cost-effective decisions. MSNN captures nonlinear demand variations, while SFO ensures efficient convergence for complex scheduling. Validation using the IEEE 24-bus system was carried out in MATLAB platform, demonstrating that proposed method outperforms established techniques as Chaos Game Optimization (CGO), Improved Particle Swarm Optimization (IPSO), and Improved Manta Ray Foraging Optimization (IMRFO), achieving 45.73% reduction in congestion management costs, prediction accuracy of 98.5%, and computation time of 3 seconds. Furthermore, it delivers the most favorable economic performance with a levelized cost of energy (LCOE) of 0.13$/kWh and net present cost (NPC) of 185,627$. Statistical results confirm significant improvements in prediction accuracy and efficiency. The proposed framework contributes to smarter, more resilient grid management under renewable uncertainty.