Research On Adaptive Recognition Of Station Facilities Service Capability Based On Video Images

At present,the scale and layout of various facilities in urban rail transit stations in China are designed statically to maximize the overall service capacity of the station.The station operation management department lacks the means to dynamically evaluate the matching degree between the service capacity of the station facilities and the uncertain demand of passenger flow.With the increasing number of railway passengers,especially in the peak period of passenger flow,the station faces a serious mismatch between the service capacity of important facilities and passenger demand.At present,the passenger flow monitoring method using AFC data(Automatic Fare Collection System)and mobile phone signaling data does not consider the imbalance of passenger demand distribution in time and space in the station,does not take into account the service perception of passengers,and the real-time data acquisition is poor.Therefore,based on deep learning theory,in this paper,a video image detection method is proposed to evaluate the adaptability of station facilities service ability and passenger flow uncertainty demand.The number of passengers queuing in the range of important facilities of the station is detected by video image,and the adaptability of station facilities service ability and passenger flow demand is evaluated by combining with the adaptive grading standard.The test results can assist managers to evaluate the operation of station facilities under dynamic passenger flow demand,and provide scientific and reasonable basis for guiding the daily operation and management of stations.Firstly,the advantages and disadvantages of the adaptive capacity method of station facilities based on pedestrian simulation system and queuing theory are analyzed through the domestic and foreign research results on the service capacity and passenger flow characteristics of station facilities.Based on the network operation management mode of the station and the wide application of video image crowd detection in the field of transportation,an adaptive evaluation and detection method of station facilities service capacity and passenger flow demand based on video image is proposed.Secondly,based on the operation management requirements and the performance requirements of the detection algorithm,combined with the research results of previous scholars,YOLO algorithm is selected as the core module algorithm in the detection process of this paper by comparing the performance of various algorithms.Thirdly,the whole process of adaptive detection algorithm based on video image recognition is constructed,including video image preprocessing,core detection module,and detection output module.Fourthly,the service capacity system of station facilities and the waiting tolerance of passengers based on queuing theory are analyzed in detail.Combined with the actual maximum service capacity of station facilities and the waiting tolerance of passengers,the grading rules of adaptability evaluation are formulated in this paper.Finally,the collected real pictures of some domestic stations are tested to verify that the method can effectively evaluate the adaptability between the service ability of important facilities in the station and the passenger flow demand,and can play a role in assisting operators to formulate operation plans and strategies.