Abnormal trajectory recognition algorithm of networked vehicles based on improved particle swarm optimization

Abstract

To improve the recognition hit rate of the abnormal trajectory of networked vehicles, reduce the actual recognition difference, and expand the corresponding trajectory coverage, this paper studies the abnormal trajectory recognition algorithm of networked vehicles based on improved particle swarm optimization. In this paper, the abnormal trajectory conversion is preprocessed, and the calculation matrix of unit recognition is designed, and then the abnormal trajectory recognition algorithm model under the improved particle swarm optimization is constructed. Finally, the design of recognition algorithm is realized by using the dimension reduction processing of abnormal data. The test results show that compared with the traditional tracking anomaly recognition algorithm test group and the traditional multi-dimensional feature anomaly recognition test group, the improved particle swarm recognition algorithm test group designed in this paper has a relatively high recognition hit rate, which indicates that the algorithm has a small error and high accuracy in the process of practical application and has practical application value.