DGSim: A Scalable Framework For Simulating Energy Consumption Of Household Appliances

Traditional household energy simulation tools often struggle to efficiently scale with increasing numbers of appliances and households, leading to high computational overhead and synchronization delays. In this work, we introduce DGSim, a scalable framework for simulating household energy consumption that supports extensive simulations with millions of instances while allowing detailed customization of appliances, usage patterns, and community demographics. DGSim incorporates stochastic techniques, parallel processing, chunking, and dynamic core allocation to improve execution efficiency. Our evaluation demonstrates that DGSim efficiently handles large workloads, utilizing configurable simulations to adapt to diverse household structures and energy consumption behaviors. By integrating flexible configurations into large-scale experiments, DGSim enables detailed analysis for demand-side management and residential energy modeling.