Font Size: a A A

Elderly's Mobility Based On Evaluation And Simulation Of High-speed Railway Transfer System

Posted on:2019-11-18Degree:DoctorType:Dissertation
Country:ChinaCandidate:W ZhaoFull Text:PDF
GTID:1362330596965935Subject:Transportation planning and management
Abstract/Summary:PDF Full Text Request
The rapid development of high-speed railway transportation network has increased the complexity of the transfer process.The railway network will cover most cities in China in 2035.However,the transport layout and service of the existing transport hub seldom consider the travel behavior characteristics of elderly passengers with special physiological and psychological conditions.As a result,the long-distance travel by elderly passengers always meet obstacles.Elderly passengers will soon be the main part of long-distance travelers.Also,they will be the main passangers of the high-speed railway by 2035.The elderly's requirement should be considered during planning,design and optimization of urban passenger transfer hubs.However,there is few study of elderly's requirement in transfer hub.Based on long-term experiments,we measure the transfer process of elderly people and the corresponding behavioral characteristics in the high speed railway station.The evaluation model and advanced simulation method of HSR transfer system is built.Firstlly,this paper constructs the experimental method of physical and psychological characteristics of general mobility for the elderly.Also we build collection model of the transfer behavior in the high-speed railway hub.Aiming at the general travel problems of the elderly,a community survey and experiment in the urban center are designed.For transferring in high-speed railway station,transfer behavior experiments and scenario experiments are designed.Which inclued the use of VR equipment.Based on this,the results of each experiment are linked with each other to form a hierarchical and cross-contrast utilization model.Which provides the data basis for the case study of the following chapters.Secondly,for daily travel,the discriminant model of elderly mobility based on the physiological and psychological characteristics is established.According to the elderly's feature,the physiological and psychological characteristics during the general travel are extracted and quantified.Advanced DNN is build to determine the ability of the elderly travel.Mathematically,we test the proposed algorithm using the actual data set collected in the previous section,then analyzes the discrimination results of the travel ability of the elderly hierarchically.Thirdly,based on the survey of Wuhan Railway Station,this paper analyzes the transfer characteristics and transfer capacity of the elderly in the HSR hub.The characteristics of transfer ability of elderly people are also presented Based on Adaptive Lasso algorithm.According to this model,the influencing factors of transfer capacity and transfer capacity are determined.Then,transfer system evaluation model based on the elderly HSR.According to the survey data and PCA,I reduce the data dimension.Then,an evaluation model of high speed railway hub transfer system based on DBN-MR is constructed.The model is validated using the statistical and machine learning perspectives.Finally,the simulation algorithm of high-speed railway hub transfer based on image recognition and case study.Using the two latest image recognition and processing techniques-an improved algorithm based on CNN and an improved TLD-,to train model and detect pedestrian flow.Based on this,a precise and innovative high-speed rail hub transfer simulation model and software for the elderly is established.Then,a case based simulation test and comparison si conducted.The results of this paper can effectively promote the development of HSR and even other large interchange hubs.The innovative experimental methods and pedestrian flow surveillance methods has been extensively tested in traffic flow and autopilot research.
Keywords/Search Tags:seniors travel, transfer hub, transfer ability, image recognition, deep learning
PDF Full Text Request
Related items