Design And Implementation Of Intelligent Track Processing System Based On Big Data

In recent years,the construction of military combat system in China is very significant,but the available means of target track intelligence analysis and prediction are relatively single,and the analysis ability is lacking.At the same time,the diversification of detection means makes the combat system accumulate a large number of basic models and military and civilian historical track data.These track data resources scattered in each acquisition system have not been fully mined and utilized.The rapid development of data mining technology provides more means for the analysis of additional information of track data.The military track data has the characteristics of purpose,regularity,region and obvious track graph.The recognition and analysis of it can effectively assist the early warning and monitoring personnel to understand the situation change more objectively and realize the intelligent processing of abnormal information.This is very important for situation assessment,threat assessment and command decision-making.Based on the above background,this dissertation designs and implements an intelligent track processing system.In the first two chapters of this dissertation,we introduce the research progress at home and abroad,Based on the related technology,an intelligent track processing system is designed to transform the abnormal behavior analysis of data mining into service.The requirements of the intelligent track processing system in this dissertation are analyzed,and the system visualization interface,track intelligent analysis,original data processing,target data storage,performance indicators are described in terms of functional requirements and non functional requirements.In the aspect of system outline design,the overall process design,module design and overall architecture design of the system are given,and then the related algorithm of pattern recognition used in intelligent analysis is given.Through the research and analysis of the method of detecting the abnormal behavior of the target track,the existing track detection mode for the abnormal position of the target is jumped out,and the track graph characteristics formed by the target track are mined.A pattern recognition algorithm based on corner detection of k-ccss(k-cosine curvature scale space)is proposed to analyze the abnormal behavior of the target track.Finally,according to the requirement analysis,the function test and non function test of the whole system are carried out,and the test results show that the platform function and performance are normal,consistent with the user requirements.Now the system has been adopted and used on line,which provides users with the auxiliary role of target track analysis,and further plays the military value of track data.