Day-ahead multi-modal demand side management in microgrid via two-stage improved ring-topology particle swarm optimization

The role of an energy management system is crucial today. However, the heavy reliance on fossil fuels as well as the growing gap between electricity demands and energy power generation have resulted in various global challenges, including energy shortages, high utility bills and greenhouse gas emissions. To address this issue, this paper presents a practical and comprehensive microgrid model that combines day-ahead multi-modal demand side management (DSM) and energy storage (ES) operation. The model aims to provide multiple optimal or nearoptimal DSM suggestions to users, increasing their willingness to respond the suggestions and fully leveraging the benefits of DSM. Besides, the ES operation works in conjunction with DSM as an energy buffer for generated power, ensuring that user demands are always met. Particle swarm optimization(PSO) is improved to optimize the DSM model due to its merits, including simplicity, population-based structure, and effective learning mechanism. To balance the two key capabilities of PSO well, the exploration ability of PSO is enhanced by indexbased ring topology, ensuring that the entire particle swarm can evenly diverge across the search space, while the exploitation ability of each sub-swarm is improved using a greedy search strategy, empowering each sub-swarm can effectively exploits their respective surroundings. To demonstrate the effectiveness of the proposed model, four different DSM strategies are designed in MG system for different purposes of cost saving, carbon emissions reduction, cost saving with load fluctuations stabilization, and emissions reduction with load fluctuation stabilization, respectively. Numerical experiments reveal advantages of the proposed ir3PSO that can search for qualified solutions with better diversity and higher accuracy in a single run. In addition, detailed sensitivities of key parameters, including load participation level and user acceptance, are also analyzed for reference by decision-makers.