巡逻式多电子哨兵目标识别的数据融合方法

对巡逻式电子哨兵目标观测在数据层的精确定位方法，以及决策层目标精确识别的数据融合问题进行研究。将个体观测的局部坐标系的目标位置转换到全局坐标系中，利用卡尔曼方法融合数据层信息。决策层目标识别的信息融合采用改进的多层次D-S证据论融合方法，将单个哨兵各异类传感器信息融合后再进行多个电子哨兵识别结果的信息融合。实验结果表明，融合后的数据稳定性和准确性都能得到提高，目标识别的正确率提高了20%。

Data Fusion Method for Target Recognition of Patrol Multiple Electronic Guards

This paper researches the precise location method for patrolling electronic guards observing object in data layer and precise identification in decision layer data fusion problem. Individual observed object data in local coordinate is transferred into global coordinate system, and Kalman method is used to realize data fusion in data layer. Improvement multilevel D-S evidence theory fusion is used in decision layer data fusion, single guard different sensor information is fused, and then multiple electronic guards observed result sensor information is fused. Experimental results show that stability and reliability of data are improved and precision of target recognition is improved by 20%.