Vehicle Collision Detection And Data Fusion Method

Intelligent Transportation Systems (ITS) can effectively improve the efficiency ofurban transport, and reduce many traffic safety problems brought by the rapid growth ofcars. Although there is no complete ITS system, the urban traffic control system hasbeen used extensively, and has become an important part of people’s life. The vehiclecollision avoidance system, which is a subsystem of ITS, is expected to reduce the lossgreatly caused by the traffic accident. As a result, it becomes a new hotspot gradually.It is the first link of the vehicle avoidance system to use sensors to detect themovement state of the preceding car and get the relevant data of the preceding car. Nowthe vehicle collision avoidance system always uses single kind of sensors to detect thetarget, but it can not explore the target accurately and efficiently because of thehardware limit of a single sensor. The data fusion of multi-sensor can integrate theadvantages of sensors and make up the disadvantages; this improves the veracity andreliability of the system. As a result，the technology of multi-sensor data fusion is moreand more popular in recent years. The main content of this paper is to use multi-sensordata fusion in the detection of vehicle collision avoidance, and to discuss how to get thedata of each sensor combined effectively and accurately. The main content is as follows.First of all, there is a lot of method about data fusion. This paper fuses the data ofinfrared sensors and radar by using the least square method and the particle filter. Atfirst the data of infrared sensors and radar is preprocessed to make them synchronized intime, and then the data of the two sensors is fused by using the least square method andweighted average, at last the data is estimated with the particle filter. In this paper thecontrast of the Kalman filter and the particle filter is simulated，which reflects thesuperiority and wide applicability of the particle filter.And then, each vehicle companies have different algorithms about how toprewarning by using the vehicle collision avoidance system. Every one hascharacteristics, but also has drawback. This paper designed a modified algorithm. itmakes the prewarning presenting to the driver visually and continually，which can makethe driver know the situation of the two vehicles directly. it also considers the influence of bad weather.