Research And Application Of Intelligent Radar Source Sorting And Threat Level Assessment

As the current electromagnetic environment becomes more and more complex,more complex radar equipments are used in large numbers.The number of radar radiation sources increases so that the pulse density increases dramatically and the pulse overlap becomes more serious.The traditional radar sorting methods can no longer meet the current demand,so it is necessary to study the radar sorting method with higher accuracy and generalizability.At the same time,in order to accurately judge the electronic battlefield situation,it is necessary to establish a scientific and reasonable radar radiation source threat level assessment model to provide accurate radiation source threat information as the theoretical basis for battlefield situation judgment.In this thesis,we firstly study the intelligent signal sorting of radar radiation sources,and analyze the traditional Pulse Repetitive Interval(PRI)algorithms and clustering-based intelligent signal sorting algorithms.Based on the research,an improved Self Adaptive Mean Shift(SAMS)signal sorting algorithm is proposed to improve the accuracy and generalizability of the signal sorting task.Through two sets of comparative experiments,it is demonstrated that the conventional PRI-based signal sorting algorithm cannot effectively sort the complex heavy-frequency radar radiation sources.In the same complex radar radiation source scenario,the Fuzzy C-means clustering algorithm(FCM),the Affinity Propagation clustering algorithm(AP),the Mean Shift algorithm(MS)and the improved SAMS algorithm studied in this thesis accurately achieved the radar signal sorting.The accuracy rates reached 87.5%,93.2%,91.8%and 93.5%.The improved SAMS algorithm optimizes the bandwidth rationality based on the MS algorithm and improves the binning accuracy and convergence speed of the MS.Compared with the FCM and AP clustering algorithms,the improved SAMS algorithm reduces the need for radar a priori knowledge and can identify arbitrary radar signal distribution shapes,which improves the generalizability and adaptability of the algorithm.After the binning is completed,based on the radar source signal binning results,the binned source parameters are evaluated in multiple layers by studying the threat level assessment method of the source.In this thesis,we proposed a threat level assessment model combining Analytic Hierarchy Process(AHP)and Intuitionistic Fuzzy Set(IFS),and introduce the Intuitionistic Fuzzy idea to optimize the weight calculation method of assessment factors.Based on the analysis of radiation source parameters and data research,the affiliation function of the weight of assessment factors is designed so that the model can reasonably assess the quantitative results of the current threat level of radiation sources.The feasibility and rationality of the model are verified through simulation experiments,and the model can provide an intuitive theoretical support for radiation source threat judgment of electronic battlefield situation.