Research On Intersection Traffic Prediction And Cab Route Recommendation Algorithm Based On Spatio-temporal Data

Traffic congestion is a worldwide problem that affects residents’ travel,reduces cab revenue and limits urban development.It is an important and challenging task in building a smart city to explore the potential value and change pattern of traffic spatio-temporal data,to provide reliable reference data for passenger search route recommendation,congestion prevention and traffic management,and to improve the probability and revenue of empty cab search.Based on the above background,this paper proposes an urban road network intersection traffic prediction model and an empty cab finding route recommendation model based on statistical indicators,neural network algorithms and spatial search algorithms,and implements a visual analysis system using visual analysis technology and intelligent interaction technology to better explore the spatial and temporal characteristics of urban intersection traffic,accurately predict urban road network intersection traffic,recommend empty cab finding routes,to assist researchers in understanding and making decisions on congestion conditions,route selection,traffic management,etc.The main work of this paper includes:1)This paper extracts intersection traffic data of urban road network based on cab trajectory data,and introduces Pearson correlation coefficients to reduce the effect of temporal similarity of intersections that are spatially adjacent but have similar temporal changes to improve Prophet temporal algorithm,and constructs intersection traffic prediction model by combining graph convolutional neural network to achieve accurate capture and prediction of intersection traffic spatial and temporal characteristics for urban traffic management and congestion prevention.It provides some reference information for urban traffic management and congestion prevention.2)The probability of picking up passengers along the route is one of the factors that cab drivers consider when choosing a route to find passengers,but this factor is ignored in the study of empty cab route recommendation.In this paper,we define the probability of picking up passengers along the route,the shortest elapsed time,the degree of traffic congestion,and the probability of picking up passengers in popular areas as cab route recommendation indicators and their calculation methods,calculate the route recommendation coefficients based on the four recommendation indicators,and build a cab route recommendation model.3)This paper designs and implements a visual analysis system for intersection flow and route recommendation,which realizes the joint and cooperative visual analysis of multiple visual views through interaction.decisions on route selection,vehicle guidance,traffic management,etc.