卷积神经网络在公交车辆调度中的应用综述OA

With the rapid advancement of deep learning,convolutional neural network(CNN)has gained widespread app-lication across various fields due to their superior performance in image processing.Leveraging its formidable capabilities in feature extraction and nonlinear modeling,CNN has gradually emerged as a pivotal technological approach for optimiz-ing urban public transportation systems.Based on existing research findings both domestically and internationally,this systematic review examines the advancements and practical applications of CNN in the field of bus vehicle scheduling.It delves into a comprehensive analysis centered around technological evolution,model architecture,and contextual applica-tions.Initially,the development history of CNN with bus vehicle scheduling is outlined,detailing their fundamental struc-ture and key operational workflows.Following this,the applications of CNN are conducted.Subsequently,a systematic review of the basic workflow of CNN is conducted,exploring pplications in urban bus scheduling,particularly in bus passenger flow prediction,arrival time prediction,and passenger crowding detection.Finally,the latest research achievements of CNN in bus vehicle scheduling are summarized and discussed,culminating in a projection of future development direc-tions to enhance the accuracy,reliability,and safety of bus scheduling.