Research On The Microscopic Behavior Models Of Vehicle, Bicycle, Pedestrian, And Their Interactive Interferences At The Signalized Plane Intersection

ABSTRACT:The traffic condition that high volumes of pedestrians and bicycles are mixed up together is one of the most obvious characteristics on the urban road network in China. In this situation, the mutual interference among pedestrians, non-motor vehicles and motor vehicles seriously affects the traffic condition of intersections; therefore, the research on the interference mechanism and providing the suitable mixed traffic microscopic behavior model become imperative, because it could not only further enrich the research in the field of mixed traffic and develop the traffic flow theory, but also guide to develop the urban traffic simulation system suitable for the specific characteristics of China, so as to provide basic parameters and theoretical support for traffic planning and management and signal control scheme.The interaction model among pedestrians, non-motor vehicles and vehicles at signalized plane intersections is proposed on the basis of extending the social force model in this dissertation, and then is exemplified by the specific study on right-turning vehicles, straight-moving bicycles from the same approach, and bidirectional pedestrians from the intersecting approach. The research process is generally as follows: microscopic behavior and traffic characteristics of right-turning vehicles, straight-moving bicycles and bidirectional pedestrians under the environment without interference from other traffic entity are analyzed, and then on the basis of the social force model, pedestrian microscopic simulation model, right-turning vehicle following model, and bicycles microscopic simulation model are built, respectively. In addition, considering that other different trip modes could influence the present users of the road, the interactive forces from those trip modes are added to the previous social force of vehicle, bicycle, and pedestrian. As a result, the behavior simulation model at the intersection where vehicle, bicycle, and pedestrian interfere with each other is finally obtained.The endeavors of this dissertation are as follows:1. Designing the improved social force model considering pedestrian deceleration avoidance. First, the pedestrian microscopic behavior and traffic characteristics are analyzed. Second, in order to introduce the "deceleration avoidance mechanism" for the social force model, the space required by pedestrian is considered, and the model gets improved. Third, the model is realized by C++programming and can simulate the phenomenon which represents that pedestrian can effectively avoid other people and obstacles and fully make use of the walk space. Through comparing the simulation data with the velocity-density diagram of pedestrian flow fitting from the real data, it shows that when the density of pedestrians is at the range of0-3.5p/m2, the proposed model can represent the basic behavior of the pedestrian flow well.2. Designing the right-turning vehicle simulation model based on the revised generalized force model and vehicle dynamic theory. First, the following behavior and traffic characteristics of right-turning vehicle are analyzed. Second, considering the difference of following behavior of right-turning vehicle and ones running on the straight road, the generalized force model gets revised. Third, a simulation frame is built, and on the basis of the revised generalized force model and Ackermann Steering Geometry, the vehicle’s accelerated velocity (velocity) and its steering angle are calculated, correspondingly. As the input of vehicle dynamic model, the steering angle and velocity are used to obtain the sideslip angle, yaw rate, and geographical coordinates when vehicle turning. At last, the data are collected, and a simulation experiment is designed, and then through comparing the flow-occupancy diagram from simulation data with that from field data, the proposed model is well verified.3. Designing the bicycle social force model considering the riding feature on the expected route. First, the microscopic behavior and traffic characteristics of bicycle are analyzed. Second, the feasibility of applying the social force model to simulate the bicycle microscopic behavior is also analyzed. Based on the above, the bicycle social force modes is constructed under the conditions that the bicycle bears three different forces, including its driving force, the outside force, and the interactive forces among its different parts. Meanwhile, based on the characteristic that bicycle runs its expected route to get through the intersection, the outside force is proposed, which includes the force from rider for driving on the expected route and the force from the barriers acting to bicycle. Due to the geometrical characteristics of bicycle, the inside interactive forces include social psychological force and retarding force. Additionally, the psychological force is used to reflect the effect of distance, and the retarding force could represent the bicycle following behavior when they run within some specific distance. Third, the data are collected, and a simulation experiment is designed, and then through comparing the velocity-density diagram from simulation data with that from field data, the proposed model is well verified.4. Designing the interactive interference model of vehicle, bicycle, and pedestrian at the signalized plane intersection. First, considering some factors, such as, the difference in velocity of vehicle, bicycle and pedestrian, different directions, and simulation step, the potential function and force of "interactive forces" of social force model are improved. The improved "interactive forces" can simultaneously consider how some factors from vehicle, bicycle and pedestrian influence itself. Those factors include the demand space when they are under next simulation step and the intersection angle among their own expected velocity. Second, through analyzing their microscopic behaviors of vehicle, bicycle and pedestrian under mixed traffic, the interactive forces from different trip modes are introduced for the above models, and then pedestrian simulation model, vehicle simulation model and bicycle simulation model are built when they are interfered by two other modes, correspondingly and respectively. Third, through comparing the tested movement trajectory data and the simulation data of pedestrian, bicycle and vehicle when they disturb anther one individually, the effect degree and scope of interactive forces from pedestrian, bicycle and vehicle are calibrated. At last, the interference model is realized on the existing mixed traffic microscopic simulation platform of signalized plane intersections, and is applied into a real case successfully.