Optimization strategy for PV-powered EV charging stations relying on V2G

<p class="p1"><span class="s1">Scientists in China have created an optimization technique for an electric vehicle charging station that uses PV, battery storage, and a vehicle-to-grid operation. Considering the uncertainty of PV generation and the randomness of EV load, the system uses two-level optimization, day-ahead and intra-day.</span></p><p>A research group from China’s <a href="https://www.pv-magazine.com/2024/06/19/chinese-researchers-develop-28-%C2%B5m-silicon-solar-cell-with-20-efficiency-0-breakage-rate/" rel="noopener" target="_blank">Shanghai Jiao Tong University</a> has developed a novel optimization strategy for an electric vehicle (EV) charging station relying on a PV system and storage solutions.</p>
<p>“This study introduces a vehicle-to-grid (V2G)-enhanced operation optimization strategy for EV charging stations with PV and energy storage (ES) integration,” said the team. “A day-ahead power purchase planning model based on two-stage distributionally robust optimization (TDRO) is established, demonstrating advantages in balancing economic efficiency with uncertainty risks. To address intra-day stochasticity, a model predictive control (MPC) based real-time optimization scheduling method for the EV charging station is proposed.”</p>
<p>The strategy was tested on a case study in Shanghai. It considers the uncertainty of PV generation and the randomness of the EV load, while applying both day-ahead and intra-day optimization.</p>
<p>The TDRO is used as a day-ahead power procurement planning model that incorporates two decision-making stages within the context of the electricity market.</p>
<p>In the first phase, the model determines the amount of electricity it needs to purchase. It uses PV forecasts and estimated time-of-use electricity pricing, while also estimating state of charge (SOC) for the batteries. In the second phase, the model considers the forecast errors of the PV system to finalize the day-ahead power procurement plan. The results are then refined on an intra-day basis.</p>
<figure class="wp-caption aligncenter" id="attachment_313952" style="width: 600px;"><img alt="" class="size-medium wp-image-313952" height="536" src="https://www.pv-magazine.com/wp-content/uploads/2025/08/1-s2.0-S0142061525005502-gr5_lrg-600x536.jpg" tabindex="0" width="600" /><figcaption class="wp-caption-text">Flowchart of the optimization strategy <p><i>Image: Shanghai Jiao Tong University, International Journal of Electrical Power & Energy Systems, CC BY 4.0</i></p>
</figcaption></figure>
<p>“The MPC rolling optimization interval is divided into two parts: one is the control interval based on short-term forecasting information, and the other is the interval based on day-ahead forecasting information,” the group explained. “Considering the balance between computational efficiency and real-time performance, and combining the practical scheduling operations of charging stations, the time scale for intra-day rolling optimization is set to 15 min. The rolling optimization interval length corresponds to the remaining time of the day. Within the MPC rolling optimization interval, decisions are made based on a combination of short-term and day-ahead forecasting information.”</p>
<p>A simulation of the optimization strategy was then carried out based on data from an EV charging station located in Shanghai, China. The station is equipped with 80 DC fast charging piles, each with a charging power of approximately 100 kW. The parking canopies feature PV panels with a total capacity of 500 kW, while two six-slot battery containers with a total capacity of 1,080 kWh are also connected on site. The EV load data was collected from July to August 2024, with a sample size of 19,570 vehicle trips.</p>
<p>The analysis showed that, compared to the original &#8220;disordered&#8221; charging, the operational costs of two typical days analyzed were reduced by 17.80% and 13.51%, respectively.</p>
<p>&#8220;Joint optimization through V2G and ES can better reduce peak loads compared to using ES alone,&#8221; the scientists concluded. &#8220;For example, with a peak load of 2,608.96 kW during the evening peak on weekdays, PV-ES optimization can reduce 11.57% of the peak load, while PV-ES-EVs optimization can achieve a 23.81% reduction.”</p>
<p>Their findings were presented in “<a href="https://www.sciencedirect.com/science/article/pii/S0142061525005502" rel="noopener" target="_blank">V2G-enhanced operation optimization strategy for EV charging station with photovoltaic and energy storage integration</a>,” published in the <em>International Journal of Electrical Power & Energy Systems</em>. Academics from China’s State Grid Shanghai Municipal Electric Power Company and <a href="https://www.pv-magazine.com/2013/08/26/goldpoly-to-acquire-400-mw-of-solar-power-assets-in-china_100012496/" rel="noopener" target="_blank">Nari Technology</a> have contributed to the study.</p>