基于径向基函数神经网络的混沌干扰信号检测  

Signal Detection in Strong Chaotic Interference Based on RBF Neural Network

作　　者：潘俊阳[1] 

机构地区：[1]西北工业大学声学工程研究所,陕西西安710072

出　　处：《计算机仿真》2010年第5期136-139,156,共5页Computer Simulation

摘　　要：为提高强混沌背景下谐波信号的检测能力,提高系统的信噪比,提出了一种在混沌背景噪声中提取正弦信号的RBF神经网络方法。依据混沌吸引子固有的几何特性和混沌系统轨迹点在流形中的演化规律,建立混沌系统的RBF神经网络单步预测模型,改进了网络的学习算法,利用RBF神经网络对输入扰动的敏感,预测出误差信号。分析了在低信噪比下的检测性能。通过对Lorenz流和实际舰船辐射噪声信号中的信号检测进行计算机仿真实验,验证了算法的有效性和可行性,并且实验表明信噪比最低达-40dB时,仍能有效检测出信号。In order to detect harmonic signal in strong chaotic interference, a new method based on RBF neural network is presented in this paper which is used to detect weak signal in strong chaotic interference. According to the geometry of chaotic attractor and the evolution of chaos trajectory, the one step prediction model of chaotic system using RBF neural network is established and a new learning algorithm is developed. By using the sensitivity of RBF neural network about input fluctuation, weak signal can be detected from the prediction error. The performance of detection for low signal - to - noise ratio （SNR） is analyzed. The validity of RBF neural network with new learning algorithm is tested by signal detection in chaotic interference in case of Lorenz flow and ship radiated noise. It had been shown that, by this method signal can be detected in chaotic interference when SNR is as low as -40dB.

关 键 词：混沌 径向基神经网络 信号检测 预测 

分 类 号：TN911[电子电信—通信与信息系统]