Influential Stations Of Urban Rail Transit Networks On Holidays And Research On Short-term Passenger Flow Forecasting

With the continuous development of the economic level and the increasingly improved urban rail transit network,the demand for transportation during holidays has increased,and the urban rail transit passenger flow is very fast to spray-type growth,which brings a huge burden on the rail network loading capacity,which will affect passenger travel efficiency And travel experience.Influential stations in the urban rail transit network not only directly affect the service level of rail transit,but also determine the efficiency of the entire rail network operation.Therefore,it is necessary to identify influential station of the track traffic network on holidays,and provide a short-term passenger flow forecast to provide scientific basis for the adjustment of operating plans during the holidays and the coordination of passenger flow.First of all,this article is classified on holidays based on the length of holiday holidays and passenger flow.Based on AFC card swiping data,analyze the characteristics and spatial distribution characteristics of different types of holiday rail stations,and provide a scientific basis for subsequent chapters.Secondly,based on complex network theory,build urban rail transit network topology models,choose Leaderank identification algorithms and improve it,comprehensively consider the centrality of the rail site,passenger flow indicators,build an influential station recognition model for the holiday urban rail transit network,and output the comprehensive importance value of each station.Influential station according to the ranking.Consider the network structure and passenger flow load capacity of the site,introduce network loss indicators and build passenger flow loss indicators,simulate deliberate attacks oninfluential stations one by one,calculate the degree of network loss and passenger flow loss after influential station fails,verify the model recognition results And compare with traditional models.Then,based on convolutional neural network infrastructure,build a holiday track station passenger flow space-time characteristic matrix,select the sparrow search algorithm,use adaptive functions to find the best parameters of its neural network,determine the optimal parameters,and build a holiday-based holiday on SSA-CNN.The short-term passenger flow prediction model of the rail station is selected as the evaluation indicator of the model prediction results.Finally,the Qingming Festival,Labor Day,National Day and Spring Festival of Xi’an Qingming Festival in Xi’an in 2021 07: 00-24: 00 time urban rail station passenger stream data as an example,output the results of influential station recognition of each holiday urban rail transit network The comparison of identification models,use the SSA-CNN prediction model to predict the passenger flow of various holidayinfluential stations in each period of time,calculate the passenger flow prediction error indicators of eachinfluential station,and compare with the CNN forecast model and SVM prediction model.Influential station recognition models and short-term passenger flow prediction models have high accuracy and strong applicability.