Scholarlink

<p>Recently, the Digital Twins (DT) have drawn the attention of the traffic community because they are thought to be one of the most effective ways to solve today's traffic issues. DT improves intelligent traffic management by simulating the transportation system, anticipating possible issues, and optimizing traffic operation. However, the complexity of contemporary urban mobility is frequently beyond the capabilities of traditional traffic management techniques. To deal with these problems, the hybrid Co-ABC optimized deep CNN is developed to create routes for vehicles and control traffic lights through a digital twin controller. The simulation includes the traffic in Dehradun city, and the resulting information is sent to the data aggregation unit in the cloud. The digital twin controller is used for effective decision-making and it contains YOLO v5, CNN, and hybrid Co-ABC optimization. YOLOv5 serves as the object prediction module and CNN is responsible for managing traffic light control by analyzing the data from YOLOv5, which includes information about the detected objects and their attributes. The sign controller operates by leveraging unique ID numbers assigned to each vehicle. This system facilitates intelligent traffic management, particularly in scenarios involving emergency vehicles. When an emergency vehicle is detected, the sign controller allocates a green light signal to the corresponding path, prioritizing the passage of the emergency vehicle. Utilizing the standard hybridization of artificial bee and coyote algorithm, the hybrid Co-ABC optimization is built for selecting the best paths and tuning the hyperparameters of CNN and YOLO v5. Subsequently, the optimized information is fed back into the vehicle network simulator, leading to improved decision-making. The proposed hybrid CO-ABC optimized deep CNN achieves impressive results, with accuracy, sensitivity, and specificity values reaching 96.74%, 96.47%, and 96.93%, for 90% of training and 95.32%, 95.54%, 95.30% for k-fold, this method demonstrates superior efficiency compared to other existing approaches.</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

Digital Twin technology for multimodal-based smart mobility using hybrid Co-ABC optimization based deep CNN

  • 期刊论文

  • 发表日期：

    2025年1月21日

查看全文
  • Mohd Anas Wajid,
  • Mohammad Saif Wajid,
  • Aasim Zafar,
  • Hugo Terashima-Marin

摘要

Recently, the Digital Twins (DT) have drawn the attention of the traffic community because they are thought to be one of the most effective ways to solve today's traffic issues. DT improves intelligent traffic management by simulating the transportation system, anticipating possible issues, and optimizing traffic operation. However, the complexity of contemporary urban mobility is frequently beyond the capabilities of traditional traffic management techniques. To deal with these problems, the hybrid Co-ABC optimized deep CNN is developed to create routes for vehicles and control traffic lights through a digital twin controller. The simulation includes the traffic in Dehradun city, and the resulting information is sent to the data aggregation unit in the cloud. The digital twin controller is used for effective decision-making and it contains YOLO v5, CNN, and hybrid Co-ABC optimization. YOLOv5 serves as the object prediction module and CNN is responsible for managing traffic light control by analyzing the data from YOLOv5, which includes information about the detected objects and their attributes. The sign controller operates by leveraging unique ID numbers assigned to each vehicle. This system facilitates intelligent traffic management, particularly in scenarios involving emergency vehicles. When an emergency vehicle is detected, the sign controller allocates a green light signal to the corresponding path, prioritizing the passage of the emergency vehicle. Utilizing the standard hybridization of artificial bee and coyote algorithm, the hybrid Co-ABC optimization is built for selecting the best paths and tuning the hyperparameters of CNN and YOLO v5. Subsequently, the optimized information is fed back into the vehicle network simulator, leading to improved decision-making. The proposed hybrid CO-ABC optimized deep CNN achieves impressive results, with accuracy, sensitivity, and specificity values reaching 96.74%, 96.47%, and 96.93%, for 90% of training and 95.32%, 95.54%, 95.30% for k-fold, this method demonstrates superior efficiency compared to other existing approaches.