大型目标红外遥测数据相关性分析与补全方法

Completion and Correlation Analysis of Infrared Telemetry Data for Large Targets

针对大型目标红外特性测试中因采用超低功耗无线遥测方案和信道劣化而导致部分测试节点的温度数据缺失问题,在考虑大型目标红外特征分布的基础上,通过分析温度数据的时间相关性和空间相关性,提出了基于单节点时间相关、多节点空间相关、环境相似条件趋势相关的数据补全算法,实现对多类缺失数据的有效补全,该算法同样适用于对测试数据中粗大误差的剔除与补全。为了验证算法有效性,通过对完整实测数据进行随机挖孔生成测试样本,将补全数据与实测数据进行比对。结果表明,补全数据的各项统计指标及与实测数据的曲线拟合度均达到较优的性能,为遥测数据的分析补全提供了一种行之有效的方案。

In terms of the loss of temperature data in some test nodes caused by adopting ultra-low power wireless telemetry and the degradation of channel, this paper puts forward the data completion algorithm related to single node time, multiple nodes space and environmental similarity conditions trend by analyzing the correlation between the time and space of temperature data on the basis of considering the infrared characteristic distribution of large targets. It effectively completes many types of data. This algorithm also applies to gross error in the data of the test strip and completion. In order to verify the effectiveness of the algorithm, the test samples were generated by random digging of the complete measured data, and the completed data were compared with the measured data. The results show that the statistical indexes of the data and the curve fitting degree of the measured data all achieve better performance, which provides an effective scheme for the analysis and completion of telemetry data.