Research On Multi-target Location And Tracking Technology Based On Anti-collosion Vehicle-Borne Radar

With the continuous development of science and technology and the continuous economic take-off in the 21st century, people's material world has been greatly enriched, car ownership has been rising year by year. In recent years, smart cars and unmanned vehicles is considered to solve the car safety and the best way to facilitate traffic and development trends.Based on this, this paper has carried on the vehicle radar localization and the track technology research. Firstly, a novel FMCW waveform is proposed for the fuzzy and false targets in multi-target recognition of vehicle-borne radar. The waveforms are limited by the velocity dimension and the distance dimension respectively, coupled with the establishment of the tolerance function, the precise number of targets can be obtained and the false targets can be effectively removed.Secondly, the characteristics and shortcomings of the existing tracking algorithms are analyzed. Aiming at the shortcomings of IMM,a new interactive multi - mode algorithm based on road information and current statistical features is proposed. The algorithm utilizes the information of the road and the current information of the vehicle. Combining the effective information can remove most of the invalid models. Only a few possible models are used to match the target's motion pattern. In this way,multiple models in the IMM algorithm will not be over-competitive, and reduce the invalid model will reduce the interference after the algorithm to improve the accuracy of filtering, but also greatly reduce the calculation.Thirdly,a new joint probabilistic data association algorithm based on dynamic threshold is proposed for the problem of data association in dense scene. The algorithm can simplify the number of correlation matrix and the number of related events for dense scene, and simplify the complexity of algorithm at the same time. Improve the convergence rate of the algorithm.Finally, the multi-radar multi-radar based multi-radar real-time scene is proposed. The single radar has the randomness and unpredictability of noise, while the multi-radar can partially cancel the noise to obtain more accurate estimation and filtering accuracy.