Scientists have developed a new optimization approach that combines both day-head optimization and real-time optimization to improve the activities of PV-driven EV charging stations. The framework is based on the auto -grazing advancing average and the Latin Hypercube sample models.
A worldwide research team led by scientists from Chinese Tianjin Renai College has developed a new stochastic optimization technique for improved shipping and operational efficiency in charging stations of PV-driven electric vehicles (EV) equipped with energy storage systems (ESS).
The proposed frameworks integrates a day -threatening optimization and real -time optimization to process forecasts, so that operational costs are minimized by dynamically planning the energy storage system.
“The increasing EV -population brings a considerable care: their random access to the grid through loading units can aggravate charging fluctuations,” the academics said. “This paper proposes a multi-Timescale stochastic Dispatch Strategy that integrates both-ahead Scheduling and intraday rolling optimization. While PV-Storage Integration and Dispatch Models Have Been Studied individually, Few Works Have-time-time-time-time Framework Under Uncertainty.
The new technology uses the AutoRegressive Moving Average (Arma) to predict the day-light PV output and EV charging tax. Subsequently, the use of Latin Hypercube sampling (LHS) makes to make many usage scenarios, so that they are later clustered in a few, to be more computational more efficiently. It uses this data to optimize the system. In the intraday level, the system uses a rolling optimization strategy on a cycle of 15 minutes, while the system is dynamically optimized.
To demonstrate the possibilities of technology, the scientists simulated three PV charging stations, each with 60 loading units of 32 kW. The PV infrastructure generates 300 kW, using 600 kW inverters with 97% efficiency and an ESS with a capacity of 800 kWh. The PV level of electricity (LCOE) was estimated at CNY 0.6041 ($ 0.085) /kWh, while the ESS CNY costs 0.3 million /kWh. The EVs have a battery capacity of 70 kWh and a loading force of 7 kW.
Four cases were tested under those circumstances; The first was a deterministic scenario without storage, and the second was a scenario with 800 kWh storage. Case three was stochast, with nine scenarios and 800 kWh storage, and case four was also stochast, with 800 kWh and 25 scenarios.
With the help of only the day draft technology, the case four had a run costs of CNY 9,716.96, case three of CNY 9,692.65, case two of CNY 9,663.15 and case one of CNY 10.916.04. With the help of the Intrada optimization, prices fell to CNY 9,283.63, CNY 9.279.53, CNY 9.274.98 and CNY 10.516.68.
“In contrast to deterministic models, which do not take prediction errors into account, and stochastic approaches with one layer that have no real-time responsiveness, the proposed method integrates both day-head and intraday optimization using a scenario-based approach, improving flexibility and accuracy,” concluded the team. “The cost savings and resilience improvements that are observed in our simulations show superior adaptability under predicted uncertainty, which underlines the benefits of our multi-Timescale stochastic optimization framework.”
The proposed methodology was presented in “Multi-TimeScal stochastic optimization for improved shipping and operational efficiency of photovoltaic charging stations of electric vehiclesPublished in the International Journal of Electrical Power & Energy Systems. Scientists from Chinese Tianjin Renai College, the Polytechnic University of Catalonia of Spain, and the Aalborg University of Denmark participated in the study.
This content is protected by copyright and may not be reused. If you want to work with us and reuse part of our content, please contact: editors@pv-magazine.com.
