CAJ | 학술논문

针对传统导弹引信不能有效识别目标的问题，提出了基于时空信息融合的复合引信目标识别方法。该方法利用BP(Back Propagation)神经网络获取基本概率赋值，根据时空信息融合算法进行证据组合，最后通过Dempster决策规则得到复合引信目标识别结果。实践表明，利用该方法进行复合引信多传感器信息融合，能够提高引信对空中目标的识别率。
침대전통도탄인신불능유효식별목표적문제，제출료기우시공신식융합적복합인신목표식별방법。해방법이용BP(Back Propagation)신경망락획취기본개솔부치，근거시공신식융합산법진행증거조합，최후통과Dempster결책규칙득도복합인신목표식별결과。실천표명，이용해방법진행복합인신다전감기신식융합，능구제고인신대공중목표적식별솔。
Considering the problem of the traditional missile fuze’s inefficacy in target recognition,a new meth-od of compound fuze’s target recognition based on temporal-spatial information fusion was put forward,which obtained the basic probability assignment based on the BP neural network,then using the temporal-spatial infor-mation fusion algorithms to assemble the evidence.In the end,the result of target recognition of compound fuze was obtained through Dempster rule.The experimental result indicated that the compound fuze could assemble the evidence of multi-sensor with this method,and enhanced the recognition of aerial-targets.