混沌时间序列中弱信号的检测

介绍了混沌的基本特性,利用混沌时间序列对RBF神经网络进行训练,则训练好的神经网络对混沌序列未来时刻的值具有一定的预测能力.若混沌序列中包含有目标信号,则其混沌特性将受到影响,导致神经网络对其预测误差的增大.这样可根据预测误差对淹没在混沌杂波及混沌杂波加一定强度的白噪声中的弱目标信号进行检测.仿真结果表明这种方法优于传统的目标检测方法,有着较好的杂波抑制能力.

Weak Signal Detection in Chaos Time Series

The basic characteristics of chaos are presented first in this paper,then the RBF neural network is trained using chaos time series and afterwards the trained neural network can be utilized to predict the chaos series of future time.The chaos characteristics are different if target signals are included in the chaos time series,which in turn will make the predicting error of the neural network become larger.Therefore,the target signals,which are submerged in chaos clutter or chaos clutter mixed with a certain intensity of white noise, can be detected on the basis of the predicting error.Simulation results indicate that this method is more effective in target detection than the conventional ones,and is more capable of suppressing clutter.