| With the dramatic increase in trip, more and more problems arise in the process of operating comprehensive transport terminal, and the available space also has a dramatic reduce. How to make the pedestrian flow in comprehensive passenger transport terminal more reasonable and effective and how to improve the space utilization have become the hot topics to discuss. And study objective for experts and scholars in this field. Many domestic scholars pay attention to the comprehensive passenger transport terminal researches,in which the pedestrian simulation research making use of computer simulation technology caused great attention. The research of pedestrian characteristics is an essential component of establishing pedestrian simulation model. Simultaneously, it is also an important basis of model calibration and simulation results evaluation. This is the significance of this paper. In another word, the research of pedestrian traffic characteristics of comprehensive passenger transport terminal is the basis of pedestrian simulation model establishment and calibration, and the basis of the evaluation of simulation results. The research of characteristics of the pedestrian has a high pertinence, because the pedestrian traffic characteristics are affected by many factors. Different scenes and different pedestrians composing will both affect pedestrian behavior in facilities,therefore, we can not blindly use the existing pedestrian models. Especially, the Compositions of Chinese transport terminals have a huge difference with those in foreign country, and the behavior of pedestrian is also different in different countries. If we use the foreign existing pedestrian model, it might affect the accuracy and reliability of pedestrian simulation. The purpose of this paper is to analyze the pedestrian characteristics of comprehensive passenger transport terminal deeply and calibrate pedestrian simulation model relying on Chinese pedestrian behavior characteristics. The research will lay a good foundation for the research work of pedestrian traffic.This paper relies on Function and Structure Experiment System of Comprehensive Passenger Transport Terminal Program, which is a Chinese National Programs of High Technology Research and Development. In July 2007, our team conducted a large-scale video data collection in Xizhimen comprehensive passenger transport terminal in Beijing. In this paper,the researches including pedestrian free speed distribution, pedestrian flow parameters relation and pedestrian micro-characteristics in comprehensive passenger transport terminals were based on the parameters of pedestrian traffic characteristics which were got from the video. The conclusions of these researches were used to calibrate parameters of pedestrian simulation model -Social Force Model, and the new parameters were verified by simulation tool.In the process of pedestrian free speed distribution research, the condition that pedestrian can walk freely was identified. And then observer chose pedestrians who matched the condition in the video as objectives to survey free speed of men and women in corridor, upstairs and downstairs in the terminal. Using statistical software SPSS to analyze speed data made the conclusion that pedestrian free speed in comprehensive passenger transport terminal matched normal distribution. The Anova analysis showed that gender and facilities type had remarkable effect on pedestrian free speed.In the process of pedestrian flow parameters relation research, there were some improvements on measurement of pedestrian flow parameters. First, when surveying the density of pedestrian flow, it should survey several times during every time-lag and then calculate the average value. Second, when surveying the speed of pedestrian flow, it should choose 25 percents of pedestrians in every time-lag and make sure they distribute averagely within time-lag to insure that survey data can represent pedestrian flow characteristics. Based on survey data, pedestrian parameter relation models in different facilities of terminal, including flow-density model, flow-space model, speed-density model and speed-space model, were established. During this process, flow-space relation curve was completed based on flow-density relation model. Meanwhile, the paper compared the study results, the minimum static space and theoretical free speed of a pedestrian acquired by these models, with those excogitated by domestic authors.In the process of pedestrian behavior characteristics research, this paper compares the pedestrian psychology decision-making process with three-level theory which was developed by S.P.Hoogendoorn to describe micro-characteristics of pedestrian behavior. The research in this part aimed at pedestrian behavior on operation level of three-level theory, and it was based mainly on a large number of observations on the behavior of pedestrians and the relative research of psychology. The main contents of the research contain pedestrian characteristics in corner, obstacles avoidance, queuing and waiting characteristics. According to the characteristics analysis, the condition in which pedestrians chose a different path was identified and so was the dynamic target selection process in the queuing area. In the process of obstacles avoidance analysis, the obstacles were classified according to the pedestrian kinematics characteristics. In the process of pedestrian waiting characteristics analysis, comparative analysis of characteristics on marked and unmarked platform was given. In the end, pedestrian size and speed in Social Force Model were acquired based on the characteristics of free-speed distribution and pedestrian flow parameters relation curves. Meanwhile, parameters of the force between pedestrians and the force pedestrian and obstacles in Social Force Model were acquired based on pedestrian flow parameters relation curves analysis combined the pedestrian behavior characteristics analysis. The calibrated parameters were verified by simulation.Concluding the research work in this paper, the main creation includes several aspects. First, this paper pointed out the path changing principle of pedestrian in turning corners, which could simulate pedestrian behaviors in the turning corners more reliably by avoiding pedestrians to congregate inboard. Secondly, time pressure was considered as a factor pattern when acquiring desired speed in Social Force Model, which could simulate the variety of desired speed. Thirdly, the space that pedestrian occupied in different conditions were analyzed based on characteristics of the pedestrian flow parameters relation curves, and then the effect distance of the force between pedestrians was calibrated, which could improve the applicability of Social Force Model to simulate interaction of pedestrian in terminals. |