Research On Multi-mode Traffic Guidance Method During Major Events Based On Group Preference

During major events,a large number of traffic and people will gather around the event area in a relatively short period of time,posing severe challenges to the road network in the area surrounding the event.How to ensure the smooth progress of major events,while ensuring smooth traffic flow during major events,and reducing the impact of events on traffic is a difficult problem to be solved.Especially in recent years,some events with major international and domestic influence have been frequently held in our country,making the research of this issue more urgent.With the development of information transmission technology,it provides technical support for traffic travel guidance.Travel guidance can greatly improve the efficiency of individual travel during major events,thereby alleviating traffic congestion in the event area.This article mainly took major events as the research object.Through the data analysis of individual characteristics and travel characteristics during the event,a model of influencing factors was constructed based on the MNL model to identify significant factors influencing the choice of travel modes for major events.Based on the density peak clustering(DPC)algorithm,the travel group was divided,the contour coefficient and the Jaccard coefficient were used to judge the clustering effect,and then the group clustering characteristic index and the group division result were determined.Based on the gray correlation model and Apriori correlation algorithm,the group travel preference was calculated,and the group travel guidance strategy was formulated according to the group preference,and the corresponding traffic travel guidance information was released for the group to realize the travel guidance based on the group preference.Finally,the evaluation index system of transportation guidance was constructed to evaluate the guidance.The main research contents of this paper were as follows:(1)Significant factors influencing the choice of individual travel mode during major events.According to the travel characteristics of major events,three new influencing factors had been added in terms of influencing factors:travel time,number of peers,and distance from the rail station.A model of influencing factors for travel mode selection based on the MNL model was constructed.The results of the model showed that educational background,monthly income,the number of disposable cars,the number of peers,travel time,and walking time from rail stations have a significant impact on the travel mode selection of individuals participating in the event.(2)Travel guidance based on group preference.Based on the density peak clustering algorithm,the value of the algorithm cutoff distance(d_c)was too dependent on subjective empirical judgment,and the introduction of the Gini coefficient(G)makes the value more objective.The gray correlation model and the Apriori correlation algorithm were used to identify group travel preferences,and finally the information needs of travel groups during major events were analyzed to lay the foundation for the release of guidance information based on travel guidance strategies.(3)Construct a travel guidance evaluation model,and build a travel guidance evaluation index system with traffic operation status,information utility,and travel satisfaction as the first-level indicators.On this basis,principal component analysis was used to determine the index weight,and then the evaluation method of extremely small index was used to divided the travel guidance evaluation into 4 levels:excellent(<0.180),good(0.180～0.281),medium(0.281～0.558),poor(>0.558).(4)Case analysis.The 18th China International Agricultural Products Fair was selected as the major activity of this article.Based on the choice of significant factors in the travel mode during major events,the improved density peak clustering algorithm was used to cluster clusters.Using the contour coefficient and Jaccard coefficient to determine the best clustering evaluation index:gender,income,education,the number of disposable cars,travel time,number of peers,the contour coefficient and Jaccard coefficient were 0.362,0.569,and finally the activity will travel Individuals were divided into 4 groups.The gray correlation model and Apriori correlation algorithm are used to identify group travel preferences.Group 1 prefered public transportation,and its correlation degree was 0.80.Group 2 prefered rail travel,and its correlation degree was 0.91.Group 3 prefered rental/online car-hailing,and its correlation degree was 0.91.Group 4 prefers car travel,and its correlation degree was 0.85.Finally,accorded to the questionnaire survey data and the traffic state data of the activity area,the evaluation level of the travel guidance program was calculated as good(0.279)used the travel guidance evaluation model proposed in this paper.