Research And Application Of Big Data Collection, Analysis And Prediction Of Intelligent Public Transport Passenger Flow

In order to respond to the call of the country to build a smart city and improve the level of digitalization and information of the Intelligent Transportation System,optimizing the mode of transportation,promoting green travel reform,and solving traffic congestion are a key task of local governments.Through real-time understanding of the changing trend of bus passenger flow,the bus capacity can be rationally dispatched,the passenger flow evacuation capacity of the station can be improved,and public transportation resources can be rejected According to the support of big passenger-flow data,it is easier to optimize the construction of bus network,improve the spatial layout of bus stations,and improve the utilization of traffic lines.This paper will focus on how to collect the number of people getting on and getting off the bus at each station in real time,then use the time series model to analyze and predict the daily passenger flow of the stationCollecting the signal of on-board automatic station announcer,analyzing arrival and departure time,station name and so on,then we can correct bus reporting error by detecting door switch and improve the detection accuracy of bus arrival and departureThrough the deep camera installed above the bus door,we collect the video of the passenger getting on and getting off the bus,and realize the passenger flow counting system The depth image is transformed into gray image,and gradient enhancement is used to sharpen the details of the image.Then the image is de-noised and filtered by OpenCV computer vision library.The roundness parameter is used to judge head feature in all feature regions extracted by MSER.By matching the head of two adjacent frames,we can recognize the trajectory of passengers and count the number of passengers getting on and getting off.In this thesis,we collect the sample data of passenger flow at each station by bus passenger flow acquisition equipment installed in Qingdao bus.The change law of passenger flow is summarized according to the time distribution characteristics of the station passenger flow.SARIMA is used to predicts passenger change after eliminating the impact of statutory holidays and emergencies,and the robustness of the prediction model can be improved.