A new intelligent charging strategy in a stationary hydrogen energy-based power plant for optimal demand side management of plug-in EVs

• Published in: International journal of hydrogen energy 2024-07, Vol.75, p.400-414
• Main Authors: Çakmak, Recep, Meral, Hasan, Bayrak, Gökay
• Format: Article
• Language: English
• Subjects: Demand side management; Hydrogen energy; Machine learning; Microgrid; PEVs; RUSBoost; Symbiotic organisms search algorithm
• ISSN: 0360-3199; 1879-3487
• DOI: 10.1016/j.ijhydene.2024.02.132

Stationary hydrogen energy-based power plants generating electricity to supply high-powered plug-in electric vehicles (PEVs) have recently become popular in renewable energy-based power plants. Besides, in a PEV charging station, various types of powered charge devices can be established such as DC fast chargers or 3.7 kW, 7.4 kW, 11 kW, and 22 kW AC chargers. This paper introduces a demand-side management-oriented optimal charging strategy that includes two stages for PEVs in a hydrogen energy-based microgrid. The paper focuses on two stages to execute an optimal charging of PEVs in compliance with their users' requests and satisfaction and considering the power system loading. It is assumed that there are three types of chargers in the PEV charging station and the users. In the first stage randomly created requests are classified by an ensemble learning classifier method that performs higher performance classification by combining the results from multiple classifiers in a machine learning classification. The second stage schedules the PEVs according to the classification results and users’ requests. To test the proposed system, first random requests are created then they are sent to the classifier, and the results of classifiers are scheduled in each other. The demand-side management-oriented charge scheduling and managing strategy which includes the proposed two stages has been compared with non-managed cases. Case study results reveal that the proposed approach provides 52.1% peak load reduction and 72.3% valley filling improvement by the SOS algorithm. The results highlight the advantages of the proposed system in terms of peak reduction and valley filling. [Display omitted] •A new intelligent decision-maker method is proposed for DSM in a microgrid.•Proposed method is developed for a stationary hydrogen energy-based power plant.•The optimal DSM of EVs is investigated with a hybrid ML method.•RUSBoost and SOS Algorithms-based Hybrid Intelligent Decision Maker is developed.•It provides 52.1% peak load reduction and 72.3% valley filling improvement.