Research On Key Technologies Of Dynamic Battlefield Environment Impact Analysis On Military Action

With the gradual development of our military data construction,the main body of military data has extended from basic attribute data to dynamic situation data,decision support data,etc.The data types and scales have grown very rapidly,forming a certain field of massive dynamic data.However,limited by the development level of the previous technology,such amount of historical accumulated data and real-time collected data are not effectively utilized.They are just used by simple storage and simple statistical analysis.Important knowledge and laws implied in these data have not been effectively explored.This paper analyzes the impact of dynamic battlefield environment on military operations by means of the latest technologies and achievements in the field of data analysis,and studies the key technologies of online mining for impact analysis on the military operation effectiveness by battlefield environment.In general,the main work of this article is as follows:1.An online analysis and assessment framework for the impact of dynamic battlefield environments on military operations is proposed.This paper summarizes the research contents of the battlefield natural environment impact analysis of military operations,establishes a framework for the analysis and assessment of the impact of dynamic battlefield environment on military operations,and sorts out relevant key technologies such as time series clustering,dynamic data stream classification,impact factor mining,fuzzy influence diagram analysis and evaluation to make further research.2.A time series clustering algorithm based on peak interval is proposed.The proposed method can effectively utilize the peak interval information inherent in the time series data set and improve the clustering performance.The experimental results on the synthetic data set and the real data set show that the proposed algorithm is effective and has higher precision than the method based on Euclidean distance and dynamic time warping.3.An online active learning ensemble framework for drifted data streams is proposed.The ensemble classifier consists of a long-term stable classifier and multiple dynamic classifiers that can efficiently process gradual drift and abrupt drift data streams.Active learning uses a non-fixed label budget policy,supporting on-demand labeling request,and adopts uncertainty strategy and random strategy to label instances.The experimental results on the synthetic data set and the real data set show that the proposed method can obtain good prediction accuracy without increasing the total cost of labeling.4.A classification impact factor mining technique based on contrast pattern is proposed.Firstly,two frequent closed itemset mining algorithms are extended to the Hadoop platform to mine closed frequent itemsets of high-speed data streams.The compressed prefix tree structure and vertical data format are adopted respectively.The experimental results show that the vertical data format can improve the processing speed.Furthermore,this paper explores the main impact factors of different data categories by mining high-quality contrast pattern equivalence classes,and theoretically analyzes the rationality and feasibility to determine classification impact factors through contrast pattern equivalence classes.The experimental results show that the proposed method is effective for mining impact factors.5.A method for analyzing the impact of battlefield environment on military operations effectiveness based on fuzzy influence diagram is proposed.Aiming at the qualitative and quantitative comprehensive analysis requirements for the impact of the battlefield environment on the military operation effectiveness,the fuzzy influence diagram evaluation method is introduced and appropriately extended.Combined with an example of anti-terrorist action,the specific process of analyzing the impact of the battlefield environment on the effectiveness of military operations based on the fuzzy influence diagram is given.In a word,this paper studies the online analysis and assessment of the impact of the dynamic battlefield environment on military operations,and establishes an online analysis framework.Research on key technologies such as time series clustering,dynamic data stream classification,impact factor mining,and fuzzy influence diagram analysis and evaluation is done.The research results are an exploration of the dynamic battlefield environment impact analysis technology for military operations,which has very realistic military value and has certain theoretical and practical significance for improving the effectiveness of military operations in complex battlefield environments.