中文摘要：

传感器在无人车间的机械手操作中发挥着重要作用。针对传感器在目标识别领域由于环境因素和自身精度限制带来的识别精度不高的问题,通过预处理和聚类,对源数据规模执行二次降维,克服证据理论的不足,提出一种基于证据理论的多传感器多目标识别方法。仿真结果表明,该方法具有较高的目标识别精度和较少的时间消耗。

英文摘要：

Sensors played significant role in the operation of manipulator in the unmanned workshop.In view of the problem that the recognition accuracy was not high due to the environmental factors and the limitation of the precision of the sensor in the field of target recognition, by preprocessing and clustering, the scale of the source data was reduced by two times, a recognition method of multi-sensor multi-target based on evidence theory was proposed overcoming the deficiency of the evidence theory.The simulation results show higher target recognition accuracy and less time consuming of this method.