Adaptive demand response-based stochastic planning and sizing of electric vehicle–integrated multi-energy microgrids

This paper presents a stochastic hierarchical planning framework for multi-energy microgrids (MGs) incorporating demand response (DR) and electric vehicle (EV) integration. The framework consists of two stages. The first stage minimizes the levelized cost of energy (LCOE) while promoting EV integration. The second stage enhances MG reliability through intra-day energy scheduling. A novel adaptive demand response (ADR) strategy is developed to schedule hourly deferrable loads over an annual planning horizon. The ADR is directly embedded within the planning framework, simplifying the problem structure while improving computational tractability. It is governed by two key parameters: the adaptation gain and a minimum threshold value defined by the system operator. In addition, a hybrid solution framework combining ADR with the Animated Oat Optimization Algorithm (ADR-AOOA) is proposed. Simulation results show that the ADR-based framework reduces costs by 5–8% and decreases CO₂ emissions by approximately 11.7–25% compared to a non-adaptive DR strategy. Under load scaling factors from 1 to 2, cost reductions of 4.5–5.8% and emission reductions of 14–18% are maintained relative to conventional DR. Sensitivity analysis indicates that LCOE is influenced by both degradation and economic uncertainties, with a 90% confidence interval ranging from 0.0831–0.086 $/kWh under degradation uncertainty and 0.0499–0.086 $/kWh under economic uncertainty. The results demonstrate that the proposed ADR-AOOA provides effective performance for the considered problem under the tested scenarios.