Robust Day-Ahead Voltage Support and Building Demand Response Scheduling Under Gaussian Mixture Model Uncertainty

The increasing penetration of distributed energy resources (DERs) and flexible loads has significantly heightened the operational complexity of active distribution networks (ADNs), especially in power and voltage control, while simultaneously creating substantial opportunities for leveraging demand-side flexibility. Although many DER management and demand response programs (DRPs) typically assume universal customer participation, their practical deployment is often hindered by technical, economic, and behavioral barriers. This paper proposes a chance-constrained, joint Volt-VAR optimization (VVO) and DRP framework specifically designed for commercial buildings (CBs) equipped with Heating, Ventilation, and Air Conditioning (HVAC) systems, electric vehicle charging stations (EVCS), battery storage, and fixed loads. The framework employs Gaussian Mixture Models (GMMs) to accurately characterize nonGaussian uncertainties in load and photovoltaic (PV) forecasts, effectively capturing rare yet impactful fluctuations. A dayahead scheduling approach simultaneously optimizes inverterbased reactive power, battery dispatch, and targeted load adjustments to minimize operational costs and ensure voltage compliance. The performance and robustness of the proposed method are validated through extensive simulations on the IEEE 123- node distribution system, demonstrating resilient scheduling even under low-probability, high-impact scenarios. By strategically targeting fewer but larger commercial loads, the approach simplifies coordination, reduces implementation barriers, and offers a scalable solution for distribution system operators aiming to effectively integrate high-value demand-side flexibility resources.