为了实现复杂战场环境下空中目标敌我属性的综合识别,在利用证据权重衡量信息可信度的基础上,提出了一种基于DS证据理论和直觉模糊集相结合(Dempster-Shafer evidence theory-intuitionistic fuzzy sets,DST-IFS)的综合敌我识别方法。首...为了实现复杂战场环境下空中目标敌我属性的综合识别,在利用证据权重衡量信息可信度的基础上,提出了一种基于DS证据理论和直觉模糊集相结合(Dempster-Shafer evidence theory-intuitionistic fuzzy sets,DST-IFS)的综合敌我识别方法。首先,分析了空中目标综合敌我识别问题,给出了具体的识别流程;然后,针对DS(DempsterShafer)函数向直觉模糊集转化过程中存在增大信息不确定性的问题,提出了一种信度分配方法用于直觉模糊隶属度和非隶属度赋值,并利用算例验证了信度分配方法的适应性和有效性;接着,给出了基于理想点法(technique for order preference by similarity to ideal solution,TOPSIS)的DST-IFS决策方法步骤;在此基础上,提出了一种基于DST-IFS的空中目标综合敌我识别的方法;最后进行了实例分析,验证了该综合敌我识别方法的有效性。展开更多
By considering distance error and direction error, Tropical Cyclone(TC) track forecasts with abnormal forecast error(AFE) at lead time of 48 h by ECMWF-IFS are selected out from 2010 to 2013. Factors closely related t...By considering distance error and direction error, Tropical Cyclone(TC) track forecasts with abnormal forecast error(AFE) at lead time of 48 h by ECMWF-IFS are selected out from 2010 to 2013. Factors closely related to AFE cases are investigated. There are 7 factors which are closely related to AFE cases. The most common one is Landfall or Passing through big island(LP) which appears 21 times among all 55 AFE cases. But LP often coexists with other factors to cause AFE cases. The second in the list is Coexistence with other TC or cloud cluster(CO) which affects more than one third of all AFE cases. Besides those 7 factors, fault of TCtracker also results in some AFE cases. There are no simple indicators for forecasters to anticipate a possible AFE case in advance. It seems that forecasters still have to anticipate AFE cases by their experiences and with synthetic analysis on all available data. Some possible ways to improve AFE cases are discussed and proposed to forecasters. That includes relying on products from ensemble prediction system or guidance from other models, simple translation process and manual analysis of TC track by forecasters under some circumstances.展开更多
文摘为了实现复杂战场环境下空中目标敌我属性的综合识别,在利用证据权重衡量信息可信度的基础上,提出了一种基于DS证据理论和直觉模糊集相结合(Dempster-Shafer evidence theory-intuitionistic fuzzy sets,DST-IFS)的综合敌我识别方法。首先,分析了空中目标综合敌我识别问题,给出了具体的识别流程;然后,针对DS(DempsterShafer)函数向直觉模糊集转化过程中存在增大信息不确定性的问题,提出了一种信度分配方法用于直觉模糊隶属度和非隶属度赋值,并利用算例验证了信度分配方法的适应性和有效性;接着,给出了基于理想点法(technique for order preference by similarity to ideal solution,TOPSIS)的DST-IFS决策方法步骤;在此基础上,提出了一种基于DST-IFS的空中目标综合敌我识别的方法;最后进行了实例分析,验证了该综合敌我识别方法的有效性。
文摘By considering distance error and direction error, Tropical Cyclone(TC) track forecasts with abnormal forecast error(AFE) at lead time of 48 h by ECMWF-IFS are selected out from 2010 to 2013. Factors closely related to AFE cases are investigated. There are 7 factors which are closely related to AFE cases. The most common one is Landfall or Passing through big island(LP) which appears 21 times among all 55 AFE cases. But LP often coexists with other factors to cause AFE cases. The second in the list is Coexistence with other TC or cloud cluster(CO) which affects more than one third of all AFE cases. Besides those 7 factors, fault of TCtracker also results in some AFE cases. There are no simple indicators for forecasters to anticipate a possible AFE case in advance. It seems that forecasters still have to anticipate AFE cases by their experiences and with synthetic analysis on all available data. Some possible ways to improve AFE cases are discussed and proposed to forecasters. That includes relying on products from ensemble prediction system or guidance from other models, simple translation process and manual analysis of TC track by forecasters under some circumstances.
