Scholarlink

The rising residential electricity demand requests for effective and responsible Demand Side Management policies. Conventional bio-inspired algorithms based schedulers are effectual but operate as black boxes, restraining user trust. This presented work introduces an Explainable AI Hybrid Scheduling technique that associate the Binary Ecological Cycle Optimizer with an explainable AI layer using SHAP Kernel. ECO optimizes electricity charge, peak-to-average ratio (PAR), and user comfort. The outcomes show a cost reduction of 17.49% and minimization of PAR to 10.3 instead of 16.39, outperforming PSO and SSA. The XAI layer explains appliances planning decisions by underlining the role of electricity prices and appliance interactions, making scheduling transparent and user friendly, and preparing the ground for credible smart grid applications.

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

XAI-HSF: An Explainable AI Hybrid Scheduling Framework for Residential Appliances in Smart Grids

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Ismael Jrhilifa,
  • Hamid Ouadi,
  • Hassan Rafia,
  • Abdelilah Jilbab,
  • Nada Mounir,
  • Youssra Lahrarti

摘要

The rising residential electricity demand requests for effective and responsible Demand Side Management policies. Conventional bio-inspired algorithms based schedulers are effectual but operate as black boxes, restraining user trust. This presented work introduces an Explainable AI Hybrid Scheduling technique that associate the Binary Ecological Cycle Optimizer with an explainable AI layer using SHAP Kernel. ECO optimizes electricity charge, peak-to-average ratio (PAR), and user comfort. The outcomes show a cost reduction of 17.49% and minimization of PAR to 10.3 instead of 16.39, outperforming PSO and SSA. The XAI layer explains appliances planning decisions by underlining the role of electricity prices and appliance interactions, making scheduling transparent and user friendly, and preparing the ground for credible smart grid applications.