无人战斗机态势评估变精度粗集决策方法

摘要

A variable precision rough set decision-making method for situation assessment of unmanned combat aerial vehicle (UCAV) based on weighted similarity measurement, which can accomplish decision-making rules matching, is put forward. A model of situation assessment decision-making information system for UCAV is established. Based on the selecting principle of β, the method of choosing value of β is presented. The deasion-making rule is then obtained from the decision-making system The method of calculating attributes' relative importance based on mutual information degree is put forward. Objective weights of attributes are presented according to the relative importance. Combined with expert experience, synthesized weights of attributes are then calculated. Finally, a weighted similarity measurement method based on the synthesized weights is developed, which could predict combat intent of one object by comparing similarity degrees between the object and other objects from the decision-making system. Simulation results show that the method is propitious to UCAV's situation assessment.%为了解决在态势评估中一般粗糙集决策方法仅对规则进行提取而不完成规则匹配的问题,提出了基于加权相似性度量的无人战斗机(unmanned combat aerial vehicle,UCAV)态势评估的变精度粗集决策方法.建立了UCAV态势评估决策信息系统模型,基于正确分类率β的确定原则,给出了β值的确定方法,在该β值下从决策模型中提取出了决策规则.提出了基于互信息度的属性相对重要度计算方法,并依据相对重要度给出了属性的客观权重,结合专家经验形成了属性的综合权重.提出了基于属性综合权重的加权相似性度量方法,通过比较当前对象与决策信息系统中所有对象的相似程度给出决策结果,完成对该对象作战意图的预测.结果表明,该方法有利于UCAV对对象作战意图的预测和判断.