Dynamic Multi-Objective Home Appliance Scheduling Using Transfer Learning Based Evolutionary Algorithm

Household Energy Management Systems (HEMS) have garnered significant public attention for their ability to optimize the operation of home appliances. Despite extensive research on optimal home appliance scheduling, these scheduling problems are inherently dynamic multi-objective optimization issues that can be effectively addressed using dynamic multi-objective evolutionary algorithms. This paper proposes a dynamic home appliance scheduling model built under time-of-use elec-tricity tariffs to minimize energy costs and user dissatisfaction, taking into account varying user priorities for each appliance at different times. A hybrid method combining elitism-based transfer learning and diversity maintenance (HETD-RM-MEDA) is utilized to solve the scheduling problem, generating optimal schedules that balance energy costs and user dissatisfaction, and providing optimal on/off states for each appliance under different environments. Simulation studies demonstrate the feasibility and effectiveness of the proposed scheduling framework for home appliances under various scenarios.