利用异类传感器的互补特性, 提出了一种新的运动目标识别算法: 选取目标速度、第1 主频、第2 主频作为识别的有效特征; 运用模糊推理得到目标分类信息的基本概率分配函数; 把D2S 证据推理和基于概率分配函数的决策相结合, 获得对目标的有效识别Z 仿真结果证明了该算法的有效性Z 关键词: 声传感器; 雷达; 目标识别; 模糊推理; D2S 证据推理 Abstract: M ak ing u se of the characterist ics of comp lem en tat ion of heterogenou s sen so rs, a new target recogn it ion algo rithm is p ropo sed. V elocity, the f irst and the second dom inan t f requency of the target are taken as the effect ive characters, and the exp ression of basic p robab ility assign2 m en t funct ion fo r the target character is ob tained th rough fuzzy inference. By com b in ing the D2S eviden t ial reason ing w ith the decision based on the basic p robab ility assignm en t funct ion, reliab le recogn it ion of the target is then ob tained. Sim u lat ion resu lt s p roved the algo rithm to be effect ive. Key words: acou st ic sen so r; radar; target recogn it ion; fuzzy inference; D2S eviden t ial reason ing