Threat estimation method based on dynamic bayesian network

Abstract

The invention discloses a threat estimation method based on a dynamic Bayes network, relating to the technical field of situation estimation. The method is realized by the following steps of 1. performing data collection and arrangement and extracting the situation elements; 2. integrating various situation elements, and establishing dynamic bayesian network topology; 3. performing network parameter learning and setting; 4. realizing fast approximation Bayesian inference based on Markov property; 5. combining each threat level probability into a continuous threat indexes and a discrete threat level; and 6. outputting the treat estimation result. The method can integrate the various situation elements, and perform reasonable and intelligent inference analysis, thereby realizing the blue group object threat dynamic quantitative and qualitative estimation and being applicable to a situation estimation and commanding control system.

Description

technical field [0001] The invention belongs to the technical field of situation estimation, in particular to a threat estimation method based on a dynamic Bayesian network, which can be used in situation estimation and command and control systems. Background technique [0002] Today's regional conflicts are characterized by diverse objects and complex environments. Faced with a sharp increase in the amount of observation data, if manual processing is still relied on, the timeliness and consistency are difficult to meet actual needs. Therefore, it is necessary to use the advantages of computer storage and computing to deal with a large number of recurring regular situations, so as to reduce the workload of the commander and enable him to grasp the real-time dynamics more quickly and effectively. Among them, the threat analysis is based on the extracted situational elements, reasoning and analysis of the threat level of the blue party in the environment, so as to provide a re...

AI Technical Summary

Problems solved by technology

The defects of the existing methods mainly include: (a) only consider the static relationship between each factor and the threat, and fail to reflect the continuous change characteristics of the threat in time; (b) estimate the threat between a single target, and the actual target is usually Tasks are executed in the form of formation group targets, the threat between individual targets does not take into account the number of targets, and it is difficult to effectively reflect the real threat situation; (c) traditional threat level estimation, usually when the probability of a certain level is greater than the threshold, will its final result

However, when the probabilities of each level are relatively close, it is difficult to set a reasonable threshold for this method, resulting in unstable output

Images