The Research Of Traffic Information Sensing Algorithm Based On Magneto Resistive Sensor

The traffic congestion has been a complex problem because of the rapid development of urbanization and the rapid growth of urban motor vehicle in recent years. Especially the vehicle density in most developed cities is centralized and huge, which cause the traffic congestion is particularly serious. A large number of practice has shown that the intelligent transportation system(ITS) can effectively alleviate traffic congestion and improve the capacity of the road network. Traffic information perception, as a key part of ITS, can provide important traffic information flow for ITS, which provides an important basis for ITS decision. Due to the advantages of small volume, low cost and high sensitivity, the magneto resistive sensor is widely used in traffic information sensing system. In this paper, geomagnetic vehicle detection system as the research object, the traffic information sensing algorithm is studied.Based on the vehicle detecting system which take the three-axis anisotropic magneto resistive sensor as a core, this paper considering the characteristics of vehicle magnetic detection signals collected by the magneto resistive sensor, especially the signals for low-speed congested traffic. A double threshold vehicle detection algorithm based on the variance of the energy signal has been proved to detect vehicles. After segment the effective vehicle signals according to the algorithm which can obtain the time of the vehicle entering and leaving the sensor detecting area, article get the traffic information about the traffic flow, lane occupancy ratio and vehicle speed.A vehicle detection algorithm based on double sensor information fusion is proposed to avoid the error of the single sensor node in the vehicle identification process. The maximum likelihood estimation method is used to fuse the vehicle signals collected by two related sensor nodes in the same network. And then the fusion of vehicle characteristic signals are taken as vehicle classification algorithm input parameters. Finally, a hierarchical decision tree algorithm is used to divide the vehicle into four categories. It contain cars, medium buses, medium trucks and buses.Ultimately, an experimental platform is built to validate the vehicle detection and vehicle classification algorithm. The experimental results have shown that the vehicle detection algorithm has high accuracy and the classification results of the classification algorithm is accurate. The algorithm has a better application in the road of traffic congestion.