摘要
提出了基于ARMA结构的自适应混响限波器(ARNAR).在频率域用短时频率方差和峰度构成起伏谱特征,并提出了基于起伏谱特征的频率归一化方法来实现恒虚警处理.最后,利用波导中回波的时间扩展现象,提出了二次快拍平均预处理的方法来实现多途情况下的CW信号检测.ARNAR与自适应陷波器(AN)相比,它具有无需知道混响的中心频率的突出优点.仿真研究表明ARNAR的止带衰减和很窄的带宽等性能略好于AN,并且带宽的控制不依赖于迭代步长的选取.对强混响背景下的低多普勒检测的仿真表明该联合检测系统是有效的.
A combined system for Doppler detection in waveguide is proposed to detect weak signals with low Doppler. A method based on adaptive noise cancelling methods using the autoregressive moving average (ARMA) was proposed to cancel the reverberation adaptively. Compared with the adaptive notch (AN) filter, it is not necessary to know the frequency of reverberation. A normalization method based on fluctuant frequency feature which was combined with frequency variance and kurtosis in frequency domain was presented to fulfill constant false alarm ratio detection. Before the constant false alarm rate (CFAR) method, making use of the echo time dispersion in shallow water, mean processing between two snap-shots was used to detect continuous wave (CW) in a multipath. The results show that frequency response of stopband filter about adaptive reverberation notch using autoregressive model (ARNAR) is better than that of AN, and the bandwidth of stopband about ARNAR is not controlled by the adaptive coefficient. The simulation experiments on detection of low Doppler signals in strong reverberation noises show that the system is effective.
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
2004年第5期574-577,586,共5页
Journal of Harbin Engineering University
基金
国家自然科学基金资助项目(60272087).
