被动传感器组网对干扰源的聚类定位算法

Clustering Algorithm of Passive Sensor Network for Locating Interference Source

为了提高对机动干扰源的定位精度,提出一种基于数据质量分析思想的多地面被动传感器组网聚类定位算法。考虑到各种客观因素和主观操作对量测的影响,该算法首先对各量测在二维空间中产生的目标估计位置点进行数据质量度量,从而确定该量测对干扰源定位结果的贡献;引入聚类思想完成对高质量位置数据的实时筛选,并由该数据实现对干扰目标的定位。通过和MMSE算法相比较,基于数据质量分析的聚类定位算法可有效消除低质量量测对定位结果的不利影响,从而提高了对干扰源的定位精度。同时对传感器布站形式的研究表明,只要对网内各传感器进行合理配置与管理,该算法即可有效降低系统偏差等不利因素对定位系统的影响。

In order to locate the maneuvering interference source more precisely, a clustering algorithm of passive sensor network for locating is proposed based on data quality analysis. Considering the effect of external factors and subjective operations on measurement, we first scale the data quality of the intersection positions formed by the given measurements, so the contribution of each measurement to the estimated interference source position can be determined. Then, clustering method is introduced for real-time filtrating of high quality intersections, and these data can be used finally to localize the interference source. By eliminating the influence of low-quality measurements to the localization, the presented algorithm can improve locating precision effectively in comparison with the MMSE algorithm. Study on sensor allocation shows that the algorithm can reduce the effect produced by system error and so on by allocating sensors rationally.