Research On Traffic Information Prediction Technology Based On Traffic Simulation

Urban traffic signal control is to improve road capacity, road traffic safety and improve the utilization of the road by regulating traffic flow the right-of-way assignment and induction. Its purpose is to improve the quality of traffic flow, better use of existing infrastructure, to achieve the purpose of alleviating traffic congestion, to improve traffic flow security,fast and comfort.At present, traffic detectors need to be installed at the intersection to collect data for traffic signal control. The purchase and maintenance of the detector takes a lot of time and money, which is a bottleneck for adaptive traffic control system, and it is difficult to achieve the effect of intelligent control greatly increased urban transportation cost.Secondly, most of the domestic urban intersection traffic control is basically a single point of control; rarely consider the correlation between the junctions, making some intersection control obvious effect, other intersection congestion deterioration. The traffic theoretically problem of the green wave have been researched, but not yet put into practice use in a large city.This paper proposed the prediction method of traffic information based on traffic simulation, if a detection device installed in the upstream section, traffic simulation method is available,while in the absence of detection device sections, apply the Calman filter prediction method, and the interaction of the two methods is added to solve the control problem of data source of traditional traffic signal control system. The effectiveness of the proposed methods are verified by real traffic data to meet the needs of future intelligent traffic control, and further to alleviate the urban traffic congestion.The main content and innovation of the paper is as follows:1. Proposed a traffic data collection idea based on traffic simulation,fusion and forecast, which provide a new concept to design and develop the future traffic control system. The world mature and commonly used traffic signal control systems need install matching traffic flow detector, while, the control design relies on the traffic flow data provided by the detector, if the detector data is not accurate or failure, that will result in controller failure, and cannot achieve the expected results; or the domestic traffic conditions are different from the control system design concept,leading the traffic signal control system cannot achieve the effect of the system desired in the country, therefore, in order to solve the problem that the detector installation is costly ,difficult to maintain, easy to damage and other issues, the idea of traffic data collection methods based on the traffic simulation online or offline simulation is proposed for controller integrated with the core model of simulation software to optimized and evaluate the control program.2. A method of traffic signal control parameter acquisition based on fusion prediction is proposed. The detection device section,using the Calman filter prediction method, and the integration of online and offline simulation results and measured data, the formation of a traffic flow data acquisition method and acquisition module integrated in the controller, the control scheme is optimized, laid the foundation for design of a new type of signal controller.3. Design a traffic signal controller framework contained simulation,fusion and forecast, mobile communication. And input parameters are provided by traffic flow detector and online simulation output data of simulation software for the design traffic signal controller. The simulation software is realized by embedded integrating in the control system as one of the data acquisition modules and to provide adequate and effective data for traffic control system. Each intersection controller was considered as a communication node with the adjacent intersection controller for communication, form a road network node communication network, and share the traffic information associated the intersections of the road network to achieve effective guidance and optimal control, and to improve road capacity and ease traffic congestions. It is a new direction for the future research of intelligent traffic control system.Finally, a case study was verified for actual road network in Beijing.It is verified that on-line or off-line simulation as one mode of control parameters input for the intelligent traffic signal controller based on simulation software is effective or not.Experimental verification of the simulation output and the actual measurement results, select the share, traffic flow, arrival time parameters are compared, it is proved that the simulation output results share and share statistical error is 0.685%, the traffic flow and the measured traffic flow error of 8.04%, time of arrival and the actual arrival time error 5.22%. The error analysis shows that, in addition to flow error is larger, because the detector detects the flow through theintersection stop line of traffic, there is no statistical time in the queue of vehicles through the intersection in the road, leading to the existence of the difference between the actual vehicle and left into the intersection stop line of vehicles. But from the error range, in the range of 10%, the accuracy of the basic guarantee in more than 90%, and the actual installation in the detector section intersection detection result is almost no difference, and the detection parameters than the actual detector is more abundant, for the proposed prediction method of traffic simulation based on traffic information to provide accurate parameters and various.Not only ensure the accuracy of the input control parameters, to provide more data to optimize the control system.