摘要根据物体的毫米波辐射特性，利用毫米波辐射计进行测量、目标识别、探测和成像等，广泛应用于环境、军事、射电天文等领域。然而通常目标的毫米波的辐射能量微弱，常被噪声所淹没，因此研究目标的弱信号探测方法是有意义的。73512

本文主要研究了通过去除接收机中的 噪声，可以减弱噪声对信号的干扰。首先，介绍了物体的毫米波辐射特性和全功率辐射计的原理。对理想的全功率辐射计进行建模，并用AR模型估计了各部分输出信号的功率谱密度。再分析了辐射计接收机系统中可能存在的噪声，对实际全功率辐射计进行建模并简化。结果证明了模型的有效性，以及用艾伦方差分析系统噪声的有效性。并用小波阈值去噪法去除 噪声，可以提高信噪比。最后对于黑体目标观测的实验数据，用艾伦方差分析了输出电压的噪声类型，再用小波阈值去噪去除 噪声，减小了输出信号受到 噪声而产生的扰动。对于成像图像，用小波阈值去噪对成像图像进行二维去噪，有效减少了图像中不必要的毛刺和斑点。结果表明，用艾伦方差分析系统噪声是可行的，以及去除接收机的 噪声对有用信号的提取是有帮助的。

毕业论文关键词  辐射特性 毫米波辐射计  艾伦方差  小波变换 去噪 

毕业设计说明书外文摘要

Title  Research on passive detection method for dim targets

Abstract According to the millimeter wave radiation characteristics of the object, using the millimeter wave radiometer to make a measure, identify target, detect target and image, etc。, it is widely used in environmental, military, radio astronomy and other fields。 However, the radiation energy of millimeter wave of target is weak, which is often submerged by noise, so it is significant to study the method of weak signal detection。

In this paper, we mainly study that the interference of noise is reduced by removing the flicker noise in the receiver。 Firstly, the millimeter wave radiometric characteristics of object and the principle of the total power radiation meter are introduced。The time and frequency model of the ideal total power radiation meter is built, and we use AR model to estimate power spectrum of output voltage of each module。 Then, the noise is analyzed that may exists in the receiver system, and the actual total power radiometer is modeled, and the model is simplified。 It is proved that the model is valid, and using the Allan variance to analysis system is effective。 The flicker noise in the receiver is removed by using the wavelet threshold de-noising method, which could improve signal to noise ratio。 Finally，for the experimental data of observing the black body, the noise type of the output voltage is analyzed with Allan variance, and then the flicker noise is removed by using the wavelet threshold de-noising method, which reduces the perturbation of the output voltage due to the flicker noise。 For the imaging, it is de-noised by using 2D wavelet threshold de-noising method, and the unnecessary rag and spot are reduced effectively。 The results show that it is feasible to use Allan variance to analyze the noise in the system, and it is helpful to get the useful signal by removing the flicker noise in the receiver。

Keywords  Radiometric character, Millimeter-wave（MMW）radiometer, Allan Variances, Wavelet transform, de-noising

目   录

1 绪论 1

1。1 前言 1

1。2 国内外研究现状 2

1。3