Scholarlink

<p>Electric Vehicles (EVs) are vehicles with rechargeable batteries that are capable of recharging from a Charging Station (CS). However, the conventional studies didn’t focus on the State-Of-Charge (SOC) of the EV battery while charging. Also, they didn’t schedule the EV regarding the peak and non-peak hours of the CS. Therefore, a framework for a multimodal EV charging system utilizing Gish activated-Long Short-Term Memory (Gish-LSTM)-centric charging requirements classification and an efficient queuing model with Circle map-based Artificial Rabbit Optimization (CARO)-centric routing is proposed, which excellently schedules the EVCS according to the peak and non-peak hours of CS. Initially, the generated solar PV power is controlled using an Extended Trapezoidal Fuzzy-Maximum Power Point Tracking (ETF-MPPT) controller. Next, the power loss that happens in the converter is reduced using Lie-Derivative Coupled Inductor Filter (Lie-D-CIF). Afterward, from the historical dataset, the week and non-week days are separated, followed by Fast Natural Visibility Graph (FNVG) generation for discovering the peak and non-peak hours. Thereafter, the historical dataset features and EV request features are given to the Gish-LSTM to identify the EV request priorities. Lastly, for locating the CS, CARO-centric routing is done regarding the EV priority, CS, and peak and nonpeak hour features. Thus, the proposed methodology performs superior to the other models in classifying the load requirements with a high accuracy (0.9892) and a minimum False Positive Rate (FPR) (0.008). Also, regarding the peak and non-peak hours, the proposed CARO selects a CS with a minimum time of 7512ms.</p> Graphical Abstract <p></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

An Effective Multimodal EV Charging Framework with Optimal Charging Station Selection By CARO and Gish-LSTM

  • 期刊论文

  • 发表日期：

    2025年10月18日

查看全文
  • Krishan Chopra,
  • Mukesh Kumar Shah,
  • K. R. Niazi

摘要

Electric Vehicles (EVs) are vehicles with rechargeable batteries that are capable of recharging from a Charging Station (CS). However, the conventional studies didn’t focus on the State-Of-Charge (SOC) of the EV battery while charging. Also, they didn’t schedule the EV regarding the peak and non-peak hours of the CS. Therefore, a framework for a multimodal EV charging system utilizing Gish activated-Long Short-Term Memory (Gish-LSTM)-centric charging requirements classification and an efficient queuing model with Circle map-based Artificial Rabbit Optimization (CARO)-centric routing is proposed, which excellently schedules the EVCS according to the peak and non-peak hours of CS. Initially, the generated solar PV power is controlled using an Extended Trapezoidal Fuzzy-Maximum Power Point Tracking (ETF-MPPT) controller. Next, the power loss that happens in the converter is reduced using Lie-Derivative Coupled Inductor Filter (Lie-D-CIF). Afterward, from the historical dataset, the week and non-week days are separated, followed by Fast Natural Visibility Graph (FNVG) generation for discovering the peak and non-peak hours. Thereafter, the historical dataset features and EV request features are given to the Gish-LSTM to identify the EV request priorities. Lastly, for locating the CS, CARO-centric routing is done regarding the EV priority, CS, and peak and nonpeak hour features. Thus, the proposed methodology performs superior to the other models in classifying the load requirements with a high accuracy (0.9892) and a minimum False Positive Rate (FPR) (0.008). Also, regarding the peak and non-peak hours, the proposed CARO selects a CS with a minimum time of 7512ms.

Graphical Abstract