A Hybrid Q-Learning Decision Support Framework for Human–AI Interactive Energy Management in Industry 5.0

Industry 5.0 revolutionizes industrial energy management by fusing human–AI collaboration within cyber-physical ecosystems to enable adaptive, resilient, and sustainable operations. This proposal introduces a Hybrid Q-learning + Rule-based Energy Management System (EMS) that unites reinforcement learning's optimization capabilities with rule-based security to harmonize distributed energy resources in a grid-connected hybrid renewable system. Experimental validation on a realistic microgrid demonstrates superior performance—Hybrid Q+Rule EMS achieves an operational profit of $0.47, while the conventional rule-based EMS incurs a $1.44 cost. The improvement stems from a 37% reduction in grid imports, $1.67 grid export revenue, and 100.4% renewable penetration. Wind power utilization rises to 89.6% (vs. 75%), with smart battery arbitrage transforming the system from cost-saving to profit-generating. This hybrid framework establishes a new paradigm for intelligent, sustainable, and economically resilient industrial energy management.