雷达数据处理与应用技术详解

4星 · 超过85%的资源需积分: 38 148 浏览量更新于2024-07-19 收藏 9.4MB PDF 举报

"雷达数据处理与应用，英文版，作者为何友、秀娟娟、关欣，由电子工业出版社于2016年首次出版，并由John Wiley & Sons Singapore Pte. Ltd.在新加坡发行，拥有全球独家授权（不包括简体和繁体中文版）以及中国大陆以外地区的电子版本非独家授权。" 《Radar Data Processing with Applications》是何友、秀娟娟和关欣合著的一本专业书籍，主要探讨了雷达数据处理的理论和技术及其实际应用。本书可能涵盖了以下几个关键知识点： 1. **雷达基础**：首先，书中可能会介绍雷达系统的基本原理，包括雷达的工作频率、波长、发射信号类型（如脉冲雷达、连续波雷达等）、天线设计以及雷达方程等基础知识。 2. **雷达信号处理**：这部分内容可能涉及信号的获取、滤波、检测和估计。书中可能详细讨论了如何对噪声背景下的雷达回波进行处理，提高信噪比，以及采用各种数字信号处理技术（如FFT、自适应滤波器等）来改善信号质量。 3. **数据采集与存储**：书中可能涵盖了雷达数据的采集系统，包括采样理论、数据压缩技术以及数据存储格式，这些都是保证数据质量和处理效率的关键。 4. **目标检测与识别**：雷达数据处理的一个重要任务是目标检测，包括单脉冲处理、动目标检测（MTI）以及利用多普勒效应进行目标分类。此外，可能还会讨论现代雷达系统中的目标识别技术，如特征提取和模式识别算法。 5. **运动目标跟踪**：书中的应用部分可能介绍了针对移动目标的跟踪算法，如卡尔曼滤波、粒子滤波等，这些算法能够持续地估算目标的位置、速度和方向。 6. **雷达系统设计与优化**：书中可能会讨论雷达系统设计的考虑因素，包括性能指标（如探测距离、角分辨率等）、功耗、成本和体积等，并探讨如何通过优化设计来提高系统性能。 7. **雷达应用实例**：除了理论知识，书中可能包含实际应用案例，比如在军事、气象、航空、交通监控等多个领域的雷达系统应用。 8. **版权信息**：此书的复制、存储和传播受到版权保护，未经许可或支付相应版税，不得以任何形式进行复制或传播。《Radar Data Processing with Applications》是一本深入雷达数据处理的专业书籍，适合对雷达系统、信号处理和相关应用感兴趣的读者，无论是学生还是从业者，都能从中获得宝贵的理论知识和实践经验。

Yang An-liang, Associate Prof. Liu Hong-ying, Lecturer Liu Hui, Lecturer Qu Lei, Lecturer

Wang Xue-sheng, Lecturer Xu Xiao-juan, Lecturer Zhang Dong-li, Lecturer Zhu Zi-jian, and

Lecturer Guan Hui-jie. The authors would like to express their appreciation to Dr. Dong Kai,

Dr. Wang Hai-peng, Dr. Cui Ya-qi, and postgraduates Miao Xu-bin, Wang Wang-song, and

Sun Shun for their participation in proofreading and revision. Special thanks go to the Electronic

Industry Publishing House, especially to Editor Qu Xin, for support in the publication of this book.

It is expected that the publication of this book will not only provide a very readable reference for

those engaged in information engineering, pattern recognition, military command, etc., but also lay

a theoretical foundation for their work and further study.

Any advice and suggestions from readers of this book are most welcome.

xviiPreface

Introduction

1.1 Aim and Significance of Radar Data Processing

Generally, a modern radar system consists of two important components: a signal processor and a

data processor. The signal processor is used for target detection (i.e., the suppre ssion of undesirable

signals produced by ground or sea surface clutter, meteorological factors, radio frequency interfer-

ence, noise sources, and man-made interference) [1–3]. When the video output signal, after signal

processing and constant false alarm rate (CFAR) detection fusion, exceeds a certain detection

threshold, it can be determined that a target has been discovered. Then, the discovered target signal

will be transmitted to the data recording device, where the space position, amplitude value, radial

velocity, and other characteristic parameters of the target are recorded, usually by computers. The

measurement output from the data recording device needs to be processed in the data processor,

which associates, tracks, filters, smooths, and predicts the obtained measurement data – such as

the target position (radial distance, azimuth, and pitch angle) and the motion parameters [4–6] –

for the effective suppression of random errors occurring during the measurement, estimation of

the trajectory and related motion parameters (velocity and acceleration, etc.) of the target in the con-

trol area, prediction of the target’s position at the next moment, and formation of a steady target

track, so that highly accurate real-time tracking is realized [7–9].

In terms of the level at which radar echo signals are processed, radar signal processing is usually

viewed as the primary processing of the information detected by the radar unit. It is done at each

radar station, with information obtained from the same radar and the same scanning period and dis-

tance unit, with the aim of extracting useful target information from clutter, noise, and various active

and passive jamming backgrounds. Radar data processing is usually viewed as secondary process-

ing of the radar information [10–13]. Making use of information from the same radar, but with dif-

ferent scanning periods and distance units, it can be done both at each independent radar station and

at the information processing center or system command center of the radar network. Data fusion of

multiple radars can be viewed as a third or tertiary processing of the radar information, which is

usually done at the information processing center. Specifically, the information the processing

Radar Data Processing with Applications, First Edition. He You, Xiu Jianjuan, and Guan Xin.

Singapore Pte. Ltd.

center receives is the measurement from the primary processing or the track from the secondary

processing (usually called the local track) by multiple radars, and the track after fusion (called

the global track or system track). The function of the secondary processing of radar information,

based on the primary processing, is to filter and track several targets, and estimate the targets’

motion parameters and characteristic parameters. Secondary processing is done strictly after pri-

mary processing, while there is no strict time limit between secondary and tertiary processing.

The third level of processing is the expansion and extension of secondary processing, which is

mainly reflected in space and dimension.

1.2 Basic Concepts in Radar Data Processing

The input to the radar data processing unit is the measurement from the front, which is the object of

data processing, wh ile the output is the track formed after data processing is conducted. Generally,

functional modules of radar data processing include measurement pretreatment, track initiation and

termination, and data association and tracking. A wave gate must be set up between the association

and the tracking process, and their relationship is shown in the block diagram in Figure 1.1. The

content and related concepts of the functional modules of radar data processing are briefly discussed

as follows.

1.2.1 Measurements

Measurements, also called observations, refer to noise-corrupted observations related to the state of

a target [14]. The measurements are not usually raw data points, but the output from the data record-

ing device after signal processing . Measurements can be divided, according to whether they are

associated with the known target track, into free measurements and correlated measurements. Free

measurements are spots that are not correlated with the known target track, while correlated meas-

urements are spots that are correlated with the known target track.

1.2.2 Measurement Preprocessing

Although modern radar adopts many signal processing technologies, there will always be a small

proportion of clutter/interference signals left out. To relieve the computers doing the follow-up

Signal

processing

Measurement

preprocessing

Data

association

Tracking Track

Gate

Track

initiation and

termination

Figure 1.1 Radar data processing relation diagram

2 Radar Data Processing with Applications

processing job from a heavy burden, prevent computers from saturation, and improve system per-

formance, the measurement given by the primary processing needs to be preprocessed, which is

called “measurement preprocessing ”: the preprocessing of secondary processing of radar informa-

tion. The preprocessing is a precondition of correct processing of radar data, since an effective meas-

urement data processing method can actually help yield twice the result with half the effort, with the

target tracking accuracy improved while the computational complexity of the target tracking is

reduced. The measurement preprocessing technology mainly involves system error registration,

time synchronization, space alignment, outlier rejectio n, and saturation prevention.

1.2.2.1 System Error Registration

The measurement data from radars contains two types of error. One is random error, resulting from the

interior noise of the measurement system. Random error may vary with each measurement, and may

be eliminated to some extent by increasing the frequency of measurement and minimizing its variance

in the statistical sense by means of methods like filtering. The other is system error, resulting from

measurement environments, antennas, servo systems, and such non-calibration factors in the data

correction process as the position error of radar stations and the zero deviation of altimeters. System

error is complex, slowly varying, and non-random, and can be viewed as an unknown variable in a

relatively long period of time. As indicated by the findings in Ref. [15], when the ratio of system errors

to random errors is greater than or equal to 1, the effect of distributed track fusion and centralized

measurement fusion deteriorates markedly, and at this point system errors must be corrected.

1.2.2.2 Time Synchronization

Owing to the possible difference in each radar’ s power-on time and sampling rate, the target meas-

urement data recorded by data recording devices may be asynchronous. Therefore, these observa-

tion data must be synchronized in multiple-radar data processing. Usually, the sampling moment of

a radar is set as the benchmark for the time of other radars.

1.2.2.3 Space Alignment

Space alignment is the process of unifying the coordinate origin, coordinate axis direction, etc. of

the data from the radar stations in different places, so as to bring the measurement data from several

radars into a unified reference framework, paving the way for the follow-up radar data processing.

1.2.2.4 Outlier Rejection

Outlier rejection is the process of removing the obviously abnormal values from radar measure-

ment data.

1.2.2.5 Saturation Prevention

Saturation prevention mainly deals with saturation in the following two cases.

1. In the design of a data processing system, there is a limit to the number of target data. However,

in a real system, saturation occurs when the data to be processed exceed the processing capacity.

3Introduction

2. The time used to process data is limited. Saturation occurs when the number of measurements,

or batches of targets, reaches a certain extent. In this case, the processing of the data from one

observation has to be interrupted before the processor starts to deal with the next batch of data.

1.2.3 Data Association

In the single-target, clutter-free environment, where there is only one measurement in the target-

related wave gate, only tracking is involved. Under multi-target circumstances, where a single meas-

urement falls in the intersection area of several wave gates or several measurements fall in the

related wave gate of a single target, data association is involved. For instance, suppose two target

tracks have been established before the radar’s nth scanning , and two echoes are detected in the nth

scanning, are the echoes from two new targets or from the two established tracks at that time? If they

are from the two established tracks at that time, then in what way can the echoes resulting from the

two scans and the two tracks be correctly paired? The answer involves data association, the estab-

lishment of the relationship between the radar measurements at a given moment and the measure-

ments (or tracks) at other moments, to check whether these measurements originate from the

processing of the same target (or to ensure a correct process of measurement-and-track pairing).

Data association, also called “data correlation” or “measurement correlation,” is a crucial issue in

radar data processing. False data association could pair the target with a false velocity, which could

result in the collision of aircraft with air traffic control radars, or the loss of target interception with

military radars. Data association is realized through related wave gates, which exclude the true

measurements of other targets and the false measurements of noise and interference.

Generally, data association can be categorized, according to what is being associated with what,

into the following classes [16]:

1. measurement-to-measurement (track initiation);

2. measurement-to-track (track maintenance or track updating);

3. track-to-track, also called track correlation (track fusion).

1.2.4 Wave Gate

In the process of target track initiation and tracking, a wa ve gate is often used to solve data asso-

ciation problems. What then is a wave gate? How many categories is it divided into? A brief

discussion of these questions follows.

An initial wave gate is a domain centering on free measurements, used to determine the region

where the target’s observations may occur. At the track initiation stage, the initial wave gate is

normally bigger for better target acquisition.

A correlation wave gate (or tracking wave gate, validation gate) is a domain centering on the

predicted position of the tracked target, used to determine the region where the target’s observations

may occur [17].

The size of the wave gate is related to the magnitude of radar measurement error, the probability

of correct echo reception, etc. That is to say, when deciding the wave gate’s shape and size,

one should make it highly probable that the true measurement falls in the wave gate, while making

sure that there are not many unrelated measurements in the correlati on wave gate. The echo falling

in the correlation wave gate is called a candidate echo. The size of the tracking gate reflects the

error in the predicted target posit ion and velocity, which is related to the tracking method, radar

measurement error, and required correct correlation rate. The size of the correlation wave gate is

4 Radar Data Processing with Applications

剩余551页未读，继续阅读

piaoxiangdahai

粉丝: 0
资源: 5

雷达数据处理与应用技术详解

MIMO radar signal processing

MIMO Radar Signal Processing

MIMO Radar Signal Processing 英文版

【Practical Exercise】Radar Signal Processing with MATLAB

Detailed Application of Window Functions in MATLAB Signal Processing

Error Analysis and System Calibration in MATLAB Signal Processing

Parameter Estimation and System Identification in MATLAB Signal Processing

【MATLAB Signal Processing for Beginners】: A First Step for Newcomers

Signal Decomposition and Reconstruction in MATLAB: Application of EMD and PCA

YOLOv10 Application Cases: Exploring Successful Practices Across Various Domains, Inspiring ...

最新资源