This research proposes a bidirectional hyperbolic graph capsule co-attention network (BHGCCAN) integrated with a pareto-optimality-based bitterling fish optimization (POBFO) algorithm for efficient energy management in grid-supported residential systems. The BHGCCAN captures complex spatiotemporal interactions among solar, wind, battery, and electric vehicle systems, while POBFO performs multi-objective optimization for real-time energy distribution. Implemented in MATLAB, the proposed model dynamically allocates energy resources based on demand, availability, and grid status. Simulation results show a 99. 8% allocation efficiency and a minimum energy cost of 1. 5, significantly outperforming conventional methods. The approach effectively reduces grid dependency and operational costs, establishing this as a workable option for environmentally friendly household energy systems.
Joseph et al. (Mon,) studied this question.