Link Clustering Method for Causal Road Analysis under Traffic Congestion Based on Relational Weights

摘要

It relates to a road clustering method based on associated weights for analyzing traffic jam-causing roads, comprising the steps of (a) generating a connection weight using road map data, (b) generating a speed weight using moving route data; (c) adding the time axis of the connection weight and speed weight generated using the convolutional layer to the associated weight; and (d) clustering the roads with the weight up to the limit by applying the soft max function to the summed associated weight. Prepare a configuration that includes steps to analyze the cause of traffic congestion based on road characteristics and connectivity between roads to alleviate traffic congestion, adjust signals on related roads, or change the number of lanes or connections on roads to reduce traffic congestion can reduce