Traffic simulation can simulate real world scenarios.Compared with traffic studies in real road environments,researchers can effectively reduce costs by using traffic simulation methods.As traffic research becomes more accurate and detailed,traffic simulation research that can truly reflect the details of traffic behavior is required.In this paper,the vehicle behavior simulation model of signal-controlled intersection protection phase is established by analyzing the vehicle behavior in the intersection and applying the vehicle agent.Using computer technology,a local micro-simulation system is established to verify the model.Firstly,the study describes the functional requirements,basic structure and decisionmaking mechanism of vehicle agent.The environment section stores basic information such as road basic alignment,traffic management control and other vehicle agent location attributes,which is the information source of vehicle agent.The physical model of vehicle agent is established,including perception module,behavior strategy module,evaluation module and behavior execution module.Under the condition of not violating traffic rules and causing serious safety risks,the vehicle agent decides the driving scheme at the next moment with the goal of maximizing individual interests.Secondly,the study establishes vehicle simulation models under the protection phase of signal-controlled intersections based on the vehicle agent.In this paper,the influence of signal lights on vehicles is discussed respectively in two main driving states: free driving state and following driving state.For the following vehicle,the interference of the ahead vehicle to the target vehicle and the possibility of the target vehicle being truncated by the signal light are considered.And,the classical following model and lane changing model are sorted out.The traditional IDM following model is optimized to form a new vehicle following model suitable for urban signal-controlled intersections.Then,the sensitivity coefficient φ is added to the IDM model to simulate the phenomenon that the vehicle at the front of the fleet reacts more sensitively to signal lights and the front vehicle when the vehicle starts.The new lane-changing model chooses the utility selection model as the basic model,and quantifies the density of vehicles on the side lane as one of the factors affecting the current lane utility value,which is used to describe the phenomenon of lane changing on the same side of multiple consecutive vehicles at an intersection.In addition,virtual vehicles with the same attributes as the target vehicle are introduced to assist lane changing behavior decision making and behavior execution.Finally,a simple local simulation system is designed and built.By comparing the simulation results with the actual data,it is proved that the simulation model proposed in this paper can describe the driving behavior of vehicles in the intersection in detail.In addition,the simulation results obtained by using the simulation model proposed in this paper are compared with those obtained by using mature vis SIM simulation software,which proves that the model proposed in this paper has advantages in detail characterization and is more consistent with the actual data in micro simulation. |