Research On The Method Of Avoiding Urban Road Traffic Congestion Driven By Travel Plan

With the rapid development of big data,cloud computing,5G,and Beidou technology,intelligent transportation is also rapidly being constructed.How to better alleviate traffic congestion is still the key to the construction.In the context of intelligent transportation,this paper introduces the idea of travel planning in response to the problems of unknowable arrival conditions in urban road traffic and delayed push of real-time information on road conditions,and establishes the entire research process on this basis in order to achieve the road transparency in the future,so that more accurate and time-efficient traffic prediction and route guidance can be carried out.With the goal of improving the status quo of traffic congestion based on travel planning ideas,the research is carried out from the following three aspects.(1)Build and optimize the travel planning system that calculates the road conditions ahead in the future.Based on the idea of travel planning,the system takes the sub-sections as the basic unit,and calculates the future location of vehicles and the traffic flow of the section where it is located by counting the request data that arrives at different times,so as to determine the congestion.The system architecture is divided into four parts.Travelers use navigation to collect travel plan data,establish a My SQL spatio-temporal database for data mining and analysis,optimize the Bi-LSTM future short-term driving speed prediction algorithm to complete the traffic flow calculation,and finally integrate in the front-end visualization page based on Gaode map API to develop corresponding user functions and road condition functions.A prototype system is developed based on this architecture,and the multi-step improvement and optimization of the original system is added in this article,aiming to achieve more accurate and scientific travel guidance from the perspective of calculation,future,and compatibility,and help the travel path optimization in the following text.(2)Establish the Bi-LSTM machine learning model to predict the future short-term driving speed of the road segment.The calculation of traffic flow in the system is based on the relationship between speed and time in travel theory,and realized by the optimized Bi-LSTM algorithm.According to the original driving route planned by the system,the road sections are modeled separately.The core of the model is designed as multi-steps sliding input and output,and two kinds of non-shared independent hidden layer training channels about sequential and reverse.The output value is based on the linear fusion of the two-way training value,and the model parameters are based on multiple experiments to get the optimal results.From the perspective of future,the optimization algorithm fully incorporates future data features and realizes the extraction of global information features.At the same time,the LSTM algorithm is selected as the control group,and the prediction results are evaluated through multiple evaluation indicators to prove the superiority of the Bi-LSTM algorithm in speed prediction,including learning ability and prediction accuracy.(3)The improved ant colony algorithm sp-aco is used to replan the congested travel plan adopting speed equivalent optimization.This part is based on the Bi-LSTM speed prediction value of the road segment to determine the congestion.Aiming at the congested road section that the vehicle will pass or arrive in the future in the original travel plan,the improved ant colony algorithm is used to optimize the travel to avoid congestion.The parameters are tuned through experiments and three improved dimensions are verified,namely,the A* thought heuristic function,the speed equivalent initial pheromone distribution,and the dual strategy pheromone update.At the same time,the application example integrates the research results of the first three chapters to realize the dynamic re-planning of the congested path of the travel plan,which not only greatly shortens the travel time,but also effectively avoids the congested road ahead.Finally,a closed-loop study is constructed,and the SP-ACO re-planning path is compared with the re-planning path of the travel planning system.The results of the two are highly consistent,which proves the efficiency of the method.Based on the idea of travel planning,this paper carries out the research on alleviating traffic congestion,hoping that through the above three systematic research modules,it can provide travelers with smarter travel route guidance,avoid congestion and improve travel experience;and provide traffic management departments with more scientific basis for congestion management,and finally make positive and effective contributions to effectively alleviate or even completely solve the problem of urban traffic congestion.It’s a good try.67 pictures,10 tables,67 references.