Research Of Data Fusion Technology Based On Vehicle Multi-radar

With the continuous advancement of technology,all technologies have higher requirements for the accuracy of data,and the radar technology has no exception.Improving the accuracy of radar data has become an inevitable trend.One method to improve radar accuracy is to grade up the radar hardware or improve signal processing algorithm progress,but this method is time-consuming,expensive and not always successful.Another method is to establish multi-radar data fusion system.The system can increase the radar coverage area and improve the accuracy of radar data within the radar overlapped coverage.This thesis takes the network radar system on vehicle as the background,and its main work is as follows:1.This thesis introduces several general network methods of multi-radar data fusion system.According to the characteristics of the radar processor AWR1642 for vehicle application,the thesis proposed and designed a new vehicle multi-radar data fusion system.2.According to the structure of the radar system,the spatial alignment module is studied according to the vehicle background.Then,based on the characteristic of the point cloud data formed by AWR1642,dense point clustering module is proposed and designed,and three kinds of single radar dense point clustering algorithm are discussed,and the accuracy of the clustering and the time efficiency are compared.The time alignment algorithm from the execution time angle and calibration error is proposed,and the optimal algorithm applied in the vehicle radar network system is obtained.For the multi-radar sparse trace clustering module,an improved maximum and minimum distance algorithm is proposed.And compared with the other two algorithms from real-time and accuracy.Finally,the weighted average cohesion method applied for the same kind of data after clustering is discussed.At the end of each module,the simulation experiment was carried out.3.The target tracking process was researched,and the algorithm of each tracking module is discussed in the thesis.After the point condensing opeation,the tracking simulation was carried out,and the result is compared with the simulation result of the single radar tracking.The comparison results show that the vehicle-based network radar system can indeed improve the accuracy of the radar data.