Scholarlink

<p>Managing the operation of microgrids with high penetration of renewable resources requires balancing cost, emissions, and supply adequacy in an uncertain environment. This research presents an integrated framework based on agent-based intelligence and fuzzy multi-objective decision making for distributed microgrid scheduling. In this framework, the Pareto front is generated using NSGA-II and MOPSO, and the final selection is made based on fuzzy satisfaction. Supply adequacy is considered as an active component in the agent coordination process. The results from 30 independent runs show that the proposed method limits the operating cost to an average of $398.2 and emissions to 248.7&#xa0;kg, and keeps the unsupplied energy at a level close to zero. The statistical test of the difference between the results confirms its significance (p-value = 0.012). Also, the scalability analysis shows that the proposed framework has a controlled computational growth with increasing system size. The findings demonstrate the efficiency and sustainability of the proposed framework and indicate its potential applicability for adequacy-aware scheduling in future smart microgrids.</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

Distributed Microgrid Scheduling Using Agentic AI and Fuzzy Multi-Objective Dispatch Balancing Cost Carbon Emissions and Supply Adequacy

  • 期刊论文

  • 发表日期：

    2026年5月28日

查看全文
  • Zornitsa Yordanova,
  • Hamed Nozari

摘要

Managing the operation of microgrids with high penetration of renewable resources requires balancing cost, emissions, and supply adequacy in an uncertain environment. This research presents an integrated framework based on agent-based intelligence and fuzzy multi-objective decision making for distributed microgrid scheduling. In this framework, the Pareto front is generated using NSGA-II and MOPSO, and the final selection is made based on fuzzy satisfaction. Supply adequacy is considered as an active component in the agent coordination process. The results from 30 independent runs show that the proposed method limits the operating cost to an average of $398.2 and emissions to 248.7 kg, and keeps the unsupplied energy at a level close to zero. The statistical test of the difference between the results confirms its significance (p-value = 0.012). Also, the scalability analysis shows that the proposed framework has a controlled computational growth with increasing system size. The findings demonstrate the efficiency and sustainability of the proposed framework and indicate its potential applicability for adequacy-aware scheduling in future smart microgrids.