Research On Highway Traffic Flow Control And Traffic Incident Detection

In 20^th century, the development of automobile traffic brings modern highway to a blossom time. Along with scale-up of intercommunion between cities and increase in the number of automobiles, people pay much more attention to safety and smooth flow in highway. So, highway traffic flow control and incident detection become important aspects of traffic research now.On the premise of automated highway, highway control system guides traffic flow according to the real time information from surveillance system, and slowly figures out traffic jam caused by traffic incidents. Real-time and effective control of traffic flow and rapid and accurate detection of traffic incidents are able to effectively improve highway capacity and utilization. The main research work and results are as follows:1. Put forward an improved traffic flow speed controller, which selects traffic flow speed as the controlled variable and adopts the backstepping technique to control the speed. Compared with former methods, the controller has a higher efficiency, and also can automatically satisfy control conditions. The simulation results show that this controller can reduce traffic jam and help to achieve smooth traffic flow on the congested highway.2. Adopt Fuzzy ART Neural Networks to analyze traffic conditions. In this way, we can detect occur and caused congestion degree of traffic incidents. According to the request of a section, the network divides traffic conditions to n styles by setting related parameters. Output of the network—real-time traffic condition is the important information for staff to judge traffic conditions in time, and then works can take measures to rule out incident combined the information from monitor . Compared with traditional detection methods, this network system not only can quickly identify the existing traffic, but also learn new traffic conditions, so this detection system can be efficiently applied to the highway traffic incident detection.3. In the end, summarize the whole work and bring up a number of further research aspects in this area.