Research On Method For Missiles Classification Based On Satellite Observation

In the modern Battlefield,it's meaningful to identify types of incoming missile target timely and accurately,which striving for the initiative of the war.But it's still very difficult to identify the specific types of incoming missile target because of the fact that intelligence information and parts of detection data can hardly be obtained.What's more,considering the actual identification requirements,we will counter the enemy effectively if we can identify the launch of the missile target,range and the other information early.First of all,this thesis discusses and extracts the feature of the missile target.The second,a classification algorithm is proposed to classify three kinds of different range missile target.In the final,the thesis studies the classification algorithm of different types of missile target,under the condition of the fact that the missile characteristic database has been acquired.The main efforts and results of this dissertation consist of several issues:a)Analysis of missile target feature.First of all,the thesis builds a missile motion model of the boost phase,analyzes the differences among the different type of missiles.Then,according to the different needs,the thesis extracts the sequence motion features which distinguish between the different ranges of missile target,the specific state motion features which distinguish between the different types of missile target.At last,missile plume infrared characteristics are introduced,differences between different types of missile target are analyzed and infrared features are extracted.b)An structure algorithm combined HMM model with PNN model is introduced to solve the question of low rate of recognition caused by the less obvious difference among the trajectory,which integrated the capacity of time sequence processing for HMM and the Bayesian decision theory for PNN.Feasibility and effectiveness of the algorithm was verified by simulation experiments.Analysis of the result shows that its performance is remarkably improved compared to traditional HMM method,with its implementation suitable for engineering fulfillment.c)Studies on classification algorithm for different type of missile target.On the basis of that we had already acquired the missile characteristic database,the thesis uses Template Matching algorithm,Gray Relational Analysis algorithm and Fuzzy membership weighted algorithm to classify the different type of the missile target for the certainty and uncertainty missile target feature database.Moreover,taking into account the different effects of different characteristics on the recognition results,the thesis uses the weight determination method based on the entropy theory to determine the weight of each feature objectively.At last,the gray correlation algorithm is improved and the weight of each feature is added to the gray relation degree.The results show that the modified algorithm has better performance than the original algorithm.The content of the thesis can classify the information of the missile early,and has a certain reference value to the research of missile target recognition.