Extraction And Analysis Of Regional Active Patterns In Public Transport Considering Different Groups Of People

With the accumulation of smart card data and positioning data in the conventional bus system,researchers can obtain information related to travel behaviors such as bus travel pattern and travel chain.The researches on smart card data mainly focus on the application of data,as well as the distribution characteristics of passenger travel.However,there is still a lot of information in IC card data that has not been fully mined and utilized.Based on the smart card data,this paper analyzes the active characteristics of different groups of people and tries to quantify the different active patterns of the population from the regional level.The main contents of the paper include the following four aspects:(1)Cleaning,screening and correcting the smart card data of the Trondheim and Beijing.Improved algorithm based on the FCM clustering method is put forward.Considering the service scope and passenger flow of the stations,the bus stations were divided into regions,provide a research basis for subsequent analysis at the regional level.(2)By using the method of SVD to determine the optimal division of time of smart card data,and count the number of card swipes,,and calculate time spent in each region.Improve the method of determining initial dimension of NMF decomposition based on SVD.The method is to obtain the optimal NMF decomposition dimension and then perform NMF decomposition to obtain the hidden active pattern of different populations in different regions.Combining the spatial distribution of regions and the proportion of eigenvalues of various active patterns,the spatial and temporal characteristics of smart card data of different types of people are analyzed.(3)To mine and analyze the active patterns of different groups of people’s bus OD information,obtain bus OD distribution maps of different active patterns,and the regional OD pair activity characteristics of different populations are extracted from travel time and space data.(4)By using the method proposed in this paper,the paper analyzes the data of the traffic card of Beijing city,extractsand compares the active patterns of different people in different areas in different cities,and discovers the commonalities and differences among different cities.