Target Clustering And Recognition In Battlefield Situation Assessment

Situation assessment is an important step of the military intelligent decision process.And target clustering is one of the most important problems in situation assessment.The results of target clustering can determine the relationship among combat targets and provide an important basis for battlefield commanders to determine tactical strategies.In this thesis,there are introductions about the concepts,basic contents and functional models of situation assessment in detail.On this basis,the concepts and basic contents of target clustering are introduced.Besides,the status researches of situation assessment and target clustering at home and abroad are introduced and sorted out in detail.And Aiming at the characteristics of air targets clustering,the researches on target clustering in battlefield situation assessment are carried out.The methods of cluster analysis are chosen to solve the problem of target clustering,and the clustering methods are improved.The main works of this thesis are as follows:(1)On the basis of the introductions to basic concepts of clustering analysis technology and related algorithms,there are many researches of several common clustering algorithms.E.g.The Nearest Neighbor algorithm,K-means algorithm and Fuzzy C-Means algorithm.And through the simulation experiments on the public data sets,the advantages and disadvantages of these algorithms are analyzed.(2)Aiming at the characteristics of air targets clustering,the maximum and minimum distance clustering algorithm is introduced.It is used to improve the classical K-means algorithm and the classical Fuzzy C-means algorithm to overcome the defects of the unknown numbers of clustering and the dependence of clustering results on the initial centers.There are introductions about the improved ideas,mathematical models and algorithm realizations of K-means clustering algorithm and Fuzzy C-means clustering algorithm based on the maximum and minimum distance clustering algorithm in detail.And the process of solving the target clustering problems and the realization of the program are given.(3)The simulation experiments of Max Min K-means algorithm and Max Min FCM algorithm in practical applications were given for specific models.E.g.The linear model,the v-shaped model and the surrounded-type model.The experimental results show that the proposed algorithm can not only improve the clustering accuracy of spherical targets,but also be applicable to air target clustering model.The main innovations of this thesis are as follows:The K-means algorithm and Fuzzy C-means algorithm are improved.And the two improved algorithms are used to solve the problems of air targets clustering in battlefield situation assessment,which good experimental results are obtained.