The Research Of Collision Warning System Based On Android Mobile Phones

Vehicle collision warning system can effectively reduce the collision accidents and improve road traffic safety. One of the key technologies is to detect and recognize pedestrians and vehicles around effectively, while the traditional detection method is mainly based on machine vision or radar, but there are some shortcomings like a blind angle existing in the inspection, poor adaptability to bad weather and so on. Considering that it only need to detect the existence of target without identifying the specific details of motion in most cases, this paper puts forward that it can use the Android smart phone that has built-in GPS, direction sensor, speed sensor to get pedestrians and vehicles" position, speed, direction and so on, on this basis, the collision warning system based on Android smart phone is proposed.Collision warning system based on Android smart phone mainly uses the smart phones, roadside equipments and the wireless communication technology. The main idea is that pedestrians and vehicles get the information such as position, speed, direction and so on through the smart phone, then use the wireless communication technology to exchange information combining with roadside equipments, and finally on the basis of obtained information it can determine whether there is the possibility of collision or not through the early warning algorithm. And then we can decide that whether give the early warning to pedestrians and drivers or not. Although the problem of blind angle existing in the inspection has been solved, the system needs to improve the positioning accuracy before it is used in intelligent traffic system because the accuracy of the smart phones’built-in GPS is poor. Without increasing the hardware cost investment, a software correcting method which named map matching is proposed to solve the problem. The main idea is to compare the phone positioning information with the road network information of electronic map database, and then the anchor point is matched to the correct position of the right road. In the matching process, it can use the Kalman filter to obtain the optimal estimates and use the multi-sensor fusion method to improve the accuracy and reliability of the matching.In order to verify the validity of the proposed map matching algorithm based on Kalman filter, first of all, it uses Matlab simulation software to simulate and it is mainly divided into two parts:straight road and the intersection. Then in the Eclipse development environment we develop the Android software, configure the roadside equipments and build the whole collision warning system based on Android smart phone so as to conduct experimental verification. Finally, through introducing electronic map, combining with the information like position, speed and direction obtained by Android smart phone, synthetically using a variety of information to realize map matching based on Kalman filter, and comparing the matching result with original GPS trajectory, electronic map road trajectory and reference data trajectory, it verifies the effectiveness of the improved algorithm.