Preference-based interactive multi-attribute decision-making support for stochastic scheduling of virtual power plants

Virtual power plants (VPPs) can smooth out the stochastic nature of renewable energy sources (RES) by modulating multiple controllable sources. To prevent power uncertainty and complementarity of heterogeneous resources, improving the electric economy and regulation capability of VPPs plays a crucial role in the multi-objective cooperation of the power system. Thus, this paper develops a day-ahead scheduling multi-attribute decision-making support (MaDMS) model, which incorporates the distributed energy resources (DERs), battery storage, and electricity consumers into VPPs, maximizing the economic and sustainability benefits while minimizing power imbalance risks of VPPs. Secondly, to solve the MaDMS, a multi-objective group search optimizer guided by interest domain (IDG-GSOMP) is proposed to search for optimal preference solutions. Finally, the quantified index system for the scheduling capability of VPPs is constructed for the state analysis of VPPs as a performance evaluation of the MaDMS scheme. According to the IEEE 30-bus system study, the IDG-GSOMP performs well in convergence and diversity in preference interest domains. Additionally, the state assessment for VPPs has verified the effectiveness of multi-resource synergistic schemes in improving grid reliability.