Explainable AI-based facility control system for energy saving and carbon reduction
摘要
Under complex manufacturing conditions, facility operations often rely on expert experience to optimize parameter adjustments, thus demanding human effort and lacking efficiency. Explainable artificial intelligence (XAI) offers insights into AI decision-making and predictions influencing facility efficiency. With heuristic algorithms such as genetic algorithms (GA), XAI enables solving optimization problems and identifying optimal parameters. This study introduces an XAI-based facility control system (XAI-FCS) that automates defining high-efficiency conditions, thereby replacing traditional methods. XAI-FCS applies these conditions as constraints for GA optimization, ensuring efficient operations, reducing energy consumption (EC) by 30.8 %, and lowering carbon emissions. It also enables parameter adjustments aligned with production schedules, saving time and costs and enhancing production line efficiency. XAI-FCS provides a reliable, efficient solution for modern manufacturing operations, reducing reliance on manual expertise while optimizing energy and operational performance.