Source-Load Coordinated Optimization Scheduling Strategy for Distributed PV and EV Charging Piles in Distribution Network

Abstract

The large-scale integration of distributed photovoltaics and electric vehicle charging stations has intensified fluctuations in the equivalent load of distribution networks, significantly a fecting power quality and supply reliability. This paper proposes an optimal dispatch strategy for distribution networks that considers coordinated operation between generation and load. First, an equivalent load model is established to analyze the impacts of photovoltaic output and charging station loads on power balance in the distribution network. Then, aiming to minimize the sum of squared changes in the equivalent load across adjacent time intervals, a coordinated optimization dispatch model is formulated by incorporating penalties for boundary power over-limitation and voltage deviation constraints.In this model, the generation side supports voltage regulation through reactive power control of photovoltaic inverters, while the load side mitigates power fluctuations via orderly charging and discharging of charging stations. Next, a multi-time-scale rolling dispatch framework comprising day-ahead, rolling, and real-time scheduling is adopted to iteratively refine charging plans, thereby enhancing adaptability to uncertainties in photovoltaic generation and load demand. Finally, case studies compare the performance under uncontrolled charging versus the proposed strategy, evaluating metrics such as equivalent load fluctuation, voltage deviation, and charging demand satisfaction rate. Results show that the proposed strategy reduces the peak-to-valley difference of the equivalent load by 39.8%, decreases the maximum boundary power over-limit by 72.3%, lowers voltage deviation from 9.1% to 5.8% (meeting the national standard of ±7%), and maintains a charging demand satisfaction rate above 96.5%. These improvements efectively enhance both power quality and supply reliability in distribution networks.