论文部分内容阅读
卫星在绕地球运行时，其周围可能带有一个或多个系留诱饵。诱饵与卫星运行于相同的轨道，具有与卫星完全相同的运动状态。为了对抗雷达和可见光探测，通过改变诱饵表面涂层，可使其具有和卫星基本相同的雷达和可见光特性。在这种情况下，红外辐射特性成了识别卫星及其诱饵的唯一特征。文章提出了一种基于红外辐射特征的区分卫星及其系留诱饵的多传感器数据融合方法，利用地面三个观测站的三个多波段传感器提供的红外特征，采用神经网络和证据理论实现多级融合，有效地解决了噪声条件下卫星及其诱饵的识别问题。
When a satellite is orbiting the Earth, it may have one or more tethered decoys around it. The bait operates in the same orbit as the satellite and has exactly the same state of motion as the satellite. In order to combat radar and visible light detection, by changing the bait surface coating, it can have basically the same radar and visible light characteristics of the satellite. In this case, the infrared radiation characteristic has become the only feature that identifies satellites and their decoys. In this paper, a multi-sensor data fusion method based on infrared radiation characteristics is proposed, which uses the multi-sensor data provided by three multi-band sensors at three observation stations on the ground. Neural networks and evidence theory are used to realize multi-sensor data fusion. Fusion, effectively solve the problem of satellite and its bait recognition under noise conditions.