论文部分内容阅读
在基于局部检测统计量的分布式检测系统中,传统的直接求和融合准则在面对局部信噪比差异较大时,检测概率下降明显。在详细分析性能损失的原因之后,扩展出一类新融合准则。新准则通过对局部统计量的幂求和,实现对高信噪比数据的更有效利用。对一个双传感器并行分布式检测系统的仿真表明,高次幂求和在局部信噪比差异明显时,性能好于低次幂,当差异减小时则相反,低次幂求和的性能更好些。高次幂融合准则的鲁棒性更强。面对不同使用环境,应该综合考虑各局部传感器脉冲积累数目和局部信噪比的可能取值范围,从而选择合理的幂阶数进行融合。
In the distributed detection system based on local detection statistics, the traditional direct summation fusion criterion decreases obviously in the face of large difference of local signal-to-noise ratio. After analyzing the causes of performance loss in detail, a new type of fusion criterion is extended. The new guidelines enable more efficient use of high S / N data by power summation of local statistics. The simulation of a dual sensor parallel distributed detection system shows that high-power summation performs better than low-power when local difference of signal-to-noise ratio is obvious, and low power sums better when the difference decreases . The high power fusion criterion is more robust. Facing different use environments, we should consider the possible accumulation range of the number of local sensors and the local signal-to-noise ratio to select a reasonable power order for fusion.