Research On Target Recognition Method Based On MMW Radar In Traffic Scene

Intelligent Transportation System(ITS)is the development trend of today's transportation system,and it is also a real-time,accurate,and efficient integrated transportation management system.In the intelligent transportation system,Millimeter-Wave(MMW)radar is a road traffic auxiliary tool,which can obtain the speed,distance,and angle of vehicles in multiple lanes in the area illuminated by the radar.MMW radar is widely used in road speeding bayonet capture and traffic information monitor.In the MMW radar traffic scene,the short distance of the target vehicle will increase the difficulty of target recognition,which will lead to clustering and data association problems.Therefore,this article mainly innovates the clustering and data association algorithms in the MMW radar target recognition technology and combines the innovative algorithm with the actual road traffic scene to design an MMW traffic radar target recognition system.The specific work is summarized as follows:(1)To solve the problem that many traditional algorithms need to determine the number of clusters in advance,a density-based adaptive distance fuzzy(DB-ADF)clustering algorithm is proposed.Firstly,the DB-ADF clustering algorithm uses adaptive distance to cluster the radar data to obtain the initial cluster center points and the number of cluster categories.Then the DB-ADF algorithm continuous iteration of the membership matrix and cluster centers to correct the results of the first clustering.Finally,the result of clustering is obtained.Experimental results show that the algorithm has higher clustering accuracy and a better clustering effect in the application of near-distance vehicle traffic scenes.(2)To solve the problem of low accuracy of initial clustering center point selection and poor clustering effect of existing clustering algorithms in traffic scenes of nearby vehicles,an adaptive ellipse distance density peak fuzzy(AEDDPF)clustering algorithm is proposed.The AEDDPF algorithm uses adaptive ellipse distance to describe radar data.The AEDDPF algorithm quickly and accurately selects the cluster centers by introducing an exponential function and obtains the number and coordinate values of the initial cluster centers.The initialized cluster center point is used as the subsequent input condition of the algorithm,which not only reduces the number of clustering iterations of the algorithm but also makes the final clustering result of the algorithm more accurate.This article also analyzes the time complexity of the AEDDPF algorithm.Finally,it also analyzes the generalization ability of the AEDDPF clustering algorithm for other types of data.The experimental results prove that the AEDDPF clustering algorithm has a good clustering effect in the application of near-driving vehicle traffic scenes,and the accuracy rate is as high as 96%.(3)To solve the problem of a large amount of calculation and a large number of related event combinations in the calculation of related events in the joint probability data association(JPDA)algorithm,a Gustafson-Kessel fuzzy joint probability data association(GK-FJPDA)algorithm is proposed.The GK-FJPDA algorithm introduces the fuzzy set theory into the JPDA algorithm and combines the advantages of the GK and the JPDA algorithm.Without reducing the accuracy of the JPDA algorithm,the correlation matrix of the target belonging to the measurement is introduced into the multi-target data association.The simulation data is used to compare the three algorithms of nearest neighbor data association(NNDA),JPDA,and GK-FJPDA.The results show that the GK-FJPDA algorithm has higher performance.(4)In the MMW traffic radar scene,an MMW traffic radar target recognition system was designed to solve the problems of indistinguishable vehicles and inaccurate data association.The system is mainly based on the above-mentioned innovative clustering and data association algorithm to complete the requirement analysis,software design,and implementation of the system function.At the same time,the feasibility and effectiveness of the innovative algorithm applied to the actual vehicle driving environment are verified.The MMW traffic radar target recognition system can meet the real-time and accuracy requirements of users during actual road testing.In MMW traffic radar target recognition technology,the clustering and data association algorithm is the focus of this article's innovation and application.At the same time,this article combines with engineering realization to propose a target recognition system based on MMW traffic radar to verify the algorithm the practical value of the innovative algorithm.