混沌噪声背景下微弱脉冲信号的检测与恢复

发布时间：2018-08-22 10:56

【摘要】：微弱信号是传统和一般的方法所不能检测到的微弱量,所谓微弱是相对于噪声而言,不只是说信号的幅度很小,主要指的是被噪声淹没的、信噪比很低的信号。微弱信号检测是利用电子学、信息论和概率统计等方法,研究被测信号的特点,分析产生噪声的原因,检测并估计被背景噪声淹没的微弱信号。微弱信号检测是人们获取信息的重要手段,在许多领域中都有比较广泛的应用。随着科学技术的发展,对微弱信号进行检测与恢复的需求日益迫切。混沌(Chaos)是一种看似无规则的运动,在确定性非线性系统中,不需要附加任何随机因素便可出现类似随机的行为。它广泛的存在于气象、水文、通信等多个领域。随着混沌理论在各个领域的发展与应用,利用混沌理论进行微弱脉冲信号的检测与估计成为一种发展趋势。从混沌噪声背景信号中检测并恢复出淹没在其中的微弱信号,尤其是混沌噪声背景下微弱脉冲信号的检测与恢复,对信号处理在理论与实践上有重要的意义。本文中,首先基于混沌信号的短期可预测性及其对微小扰动的敏感性对观测信号进行相空间重构、建立局域线性自回归模型(Local Linear Autoregressive model,LLAR)进行单步预测,得到预测误差,并利用假设检验方法从预测误差中检测观测信号中是否含有微弱脉冲信号。然后,对微弱脉冲信号建立单点跳跃模型,并融合局域线性自回归模型,构成双局域线性模型(Double Local Linear model,DLL),以极小化DLL模型的均方预测误差为目标进行优化,采用向后拟合算法估计模型的参数,并最终恢复出混沌噪声背景下的微弱脉冲信号。最后,基于典型混沌时间序列Lorenz系统进行仿真实验,结果表明,本文所建的模型对混沌噪声背景中微弱脉冲信号检测与恢复有比较好的效果。
[Abstract]:Weak signal is a weak signal which can not be detected by traditional and general methods. The so-called weak signal is not only that the amplitude of the signal is very small, but mainly refers to the signal which is submerged by noise and has low signal-to-noise ratio (SNR). Weak signal detection is based on the methods of electronics, information theory and probability statistics to study the characteristics of the measured signal, analyze the causes of the noise, detect and estimate the weak signal submerged by the background noise. Weak signal detection is an important means for people to obtain information. It is widely used in many fields. With the development of science and technology, the need for weak signal detection and recovery is becoming more and more urgent. Chaotic (Chaos) is a seemingly irregular motion. In deterministic nonlinear systems, stochastic behavior can occur without any additional random factors. It widely exists in many fields such as meteorology, hydrology, communication and so on. With the development and application of chaos theory in various fields, the detection and estimation of weak pulse signal using chaos theory has become a development trend. The detection and recovery of weak signals submerged in chaotic noise background signal, especially the weak pulse signal under chaotic noise background, are of great significance to signal processing in theory and practice. In this paper, based on the short-term predictability of chaotic signals and their sensitivity to small perturbations, the phase space reconstruction of observed signals is carried out, and a local linear autoregressive model (Local Linear Autoregressive model LLAR) is established for single-step prediction, and the prediction errors are obtained. The hypothesis test method is used to detect whether the observation signal contains weak pulse signal from the prediction error. Then, the single point jump model of weak pulse signal is established, and the local linear autoregressive model is fused to form a bilocal linear model, (Double Local Linear model-DLL), which optimizes the mean square prediction error of the minimized DLL model. The parameters of the model are estimated by backward fitting algorithm, and the weak pulse signal under the background of chaotic noise is finally recovered. Finally, based on the typical chaotic time series Lorenz system, the simulation results show that the proposed model has a good effect on the detection and recovery of weak pulse signal in chaotic noise background.
【学位授予单位】：重庆理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN911.23

【参考文献】