Road Based On Kalman Filter Target Detection And Tracking Study

With the rapid development of expressway construction in China, the prevention to expressway accidents has been paid close attention to. Developing the rear-end collisionavoidance system is one of the efficient method to decrease rear-end collision accidents on expressway.Based on the model of rear-end collision, the algorithm of target judge effectively is introduced, and the criterion of latent threatening targets is built up, which is the theoretic. Base to decrease the nuisance alarms and improve the warning system's reliability. Adapted the millimeter wave frequency modulated pulse-Doppler radar after carrying on the compare of the distance detect techniques on hoard; Analyzing the complicated circumstance during road target track carefully, the data association in radar data process and the start and end in target track and studied, In order to detect and track forward moving target in real time, by Interacting Multiple Models algorithm, the pulse signals of Doppler Radar are filtered by Kalman filter, and the relative lateral orientation angle rate between the forwarding vehicle and the following vehicle is estimated smoothly.The simulating results show that IMM algorithm used is simple and effective. The algorithm can track various motor conditions, it has a high track accuracy during the mixed wave circumstance, so that IMM algorithm can be used to track radar targets on-board reliably in real time. A good basic study is provided for rear-end collision warning system.