基于dsp的数字车流量检测雷达关键技术研究.doc

基于DSP的数字车流量检测雷达关键技术研究

摘    要

随着智能交通系统在城市交通建设中的迅速发展，系统中的各个环节发生着越来越紧密的联系，所采用的技术也在不断更新，向着更准更快更兼容的方向进一步的拓展。作为智能交通系统中不可或缺的一环-车流量检测系统，也经历了由最初的单一的接触式测量方式到互相辅助的混合型非接触测量方式转化等一系列的变化。作为一种典型的非接触式测量器，微波车流量检测雷达具有优良的特性，使用方式简便，其应用前景十分广阔。 

本文以调频连续波体制下的雷达工作原理以及微波车流量检测雷达各项功能的应用为背景，以第一代微波车流量检测雷达功能框架为基础，通过对数字雷达技术的硬件和软件的多角度深入研究设计和改进，实现一个准确度高，功能多样化，可扩展可持续发展的全天候产业化流量检测系统，并解决各种从科研产品过渡到工业产品时所面临的问题，为系统产业化打下良好基础，满足其嵌入更广大的智能交通系统所需要的各种要求。

本文首先介绍了雷达的工作原理，并对不同的雷达工作体制的原理以及应用作了简单的介绍。详细介绍了如何根据实际系统的要求提出车流量检测雷达系统的性能技术参数化要求以及该检测系统的实际工作体制和工作频段等各项指标。

为了使整个车流量检测系统使用更加方便，符合小型化的要求，大量数字化技术的运用仍是十分理想的选择。但是在基于ARM的第一代车流量检测雷达中[1]，其信号处理部分比较繁琐，采用了ARM，DSP，FPGA等多种器件，数字板体积大，不利于与雷达前端结合，没有充分利用器件的功能。本文针对系统的实际要求，基于DSP芯片实现雷达信号处理部分所需的全部功能，详细介绍了DSP芯片的使用开发方案，提出了硬件及其驱动软件的设计方法。

另一方面，车流量检测系统的信号处理算法部分的设计也是保证系统精确度的重要因素。本文首先介绍了系统的原始算法结构，然后将整个算法分割为若干模块，针对这些不同模块提出了各种算法改进方案。首先提出了提高信号采集模块性能的若干算法改进，比如如何优化选择系统的采样频率以及归一化频谱变换在DSP中的必要性和应用方法；接下来基于现有的目标判定模块提出多次判定检测法，以便在系统零信号出现时，最大程度保证系统检测概率的准确度。最后阐述了一种能有效提高雷达测量精度的背景功率谱识别方法，提出并实现了该算法在DSP中运行的方案，详细论述它是如何有效的提高了系统的检测精度。此方法已经在车流量检测雷达中得到了应用，实验效果明显。

在大量的实地外场实验和各种可靠性测试后得到了很多有效的实验数据。实验结果较好的证明了该系统的可行性和优良性，同时通过对数据的分析和评测，也发现了系统的一些问题，综合实验数据和测试过程中遇到的各种问题，本文首先提出了基于现有芯片的几个改进方案，使用不需要CPU干涉的DMA方式产生雷达调治信号以及利用不同的实时时钟中断方式来满足系统信息传输模块的需求，另外还介绍了一种标准的Modbus 通信协议来提高串口传输数据的稳定性和准确性。然后从更换更有效的DSP芯片TMS320VC5502的角度出发，将两款芯片进行了性能上的对比以及相应的驱动更改方案，针对在新芯片的各项功能在重构该系统提出了驱动设计的主体框架结构。

关键字：车流量检测雷达，数字雷达技术，多次判定检测法，背景功率谱识别

Digital Radar Technique Research of Vehicle-Flux Detecting 

Radar Based on DSP

Zheng Chun

（Microelectronics and Solid Electronics）

Directed by: Yu Wen

Abstract

With the development of Intelligent Transportation Systems(ITS) in the construction of city transportation, each part of the systems interacts with each other much more closely than ever before. Meanwhile, the technologies applied in ITS have changed greatly, which are faster, more precise and compatible. The vehicle-flux detecting system, which is the key knot of the whole ITS chain, has been improved as well. The vehicle volume detector has been changed from the single touched loop to mixed untouched methods. As one typical kind of non-touch detecting instrument, microwave vehicle-flux detecting radar is of many excellent characteristics and has been widely used.

Based on the detecting principles and the various applications of the first generation microwave vehicle-flux detecting FMCW radar[1], this paper aims at the thorough research，design and improvement of the hardware and software of digital radar technologies. Furthermore, to lay solid foundations for the industrialization of the radar sets and satisfy sorts of needs when embedded into broader ITS, this paper designs and implemented an all-day flux-detecting system with high accuracy, multiple functions and continuous development. Also many kinds of practical problems met in the industrialization have been solved in this paper.

In the beginning, how the radar practically works is described in detail and different radar mechanisms and their applications are discussed. According to the requirements of the application environment of the system, how to fix the specific parameters and functions, such as the mechanism and the working frequency, are specifically introduced.

In order to minimize the detecting radar system and make it more convenient in use,

a lot of digital techniques are needed. As to the first generation of this radar system which is based on ARM, DSP and FPGA, we can see that it is very complicated and too big to be integrated with the front of radar. What’s more, most of the chips are not fully utilized. In this paper, based on DSP chips, a new system to implement all the required and additional functions are designed. The developing scheme of the DSP chip is discussed and the architectures of the hardware and software are proposed.

On the other side, the designs of the arithmetic of the system are also essential which can guarantee the detecting accuracy. In chapter four, the original arithmetic structure is divided into several smaller modules. As to each module, the improving methods are presented and proved by reasoning and calculating. In the signal sampling module, the optimization of the sampling frequency is realized and the normalized frequency transition is implemented in DSP. Additionally, in the target-judgment module，multi-judgment detecting method is successfully applied so as to get rid of the consequences of zero signal and make the judgment of target cars more precise. Finally, a recognizing method of background power spectrum is introduced as well as the reason why it significantly improves the detecting accuracy. Also it has been transplanted into DSP and proved effective through a large amount of experiments.

The results of field experiments and reliability tests indicate that this microwave vehicle-flux detecting radar is efficient and has good performances. In the meanwhile, through the analysis and test of the experimental datas, some disadvantages have been found. Therefore, several plans based on the existed chips to improve this system are discussed such as the generation of moduling signals by DMA, the application of RTC interrupt and the Modbus protocols in the communication module. In the end, the system is reconstructed based on TMS320VC5502, for its advantages that can’t be derived in TMS320VC5416.

Key words: vehicle-flux detecting radar, digital radar technique， Multi-judgment detecting method, recognizing method of background power spectrum

目    录

摘    要 I

Abstract III

目    录 I

第1章 绪 论 1

1.1 智能交通系统概述 1

1.2 车流量检测系统的研究现状和性能比较 1

1.2.1 车流量检测器的常用类型 2

1.2.2 微波车流量检测器的优良特性 3

1.3 数字技术在雷达系统中的应用 3

1.4 本文主要内容简介 4

第2章 车流量检测雷达系统总体设计 6

2.1 车流量检测雷达的基本工作原理 6

2.1.1 车流量检测雷达工作频段及体制 6

2.1.2 车流量检测雷达工作原理 7

2.1.3 车流量检测雷达的功能 8

2.2 基于DSP的车流量检测雷达系统设计方案 9

2.2.1 系统硬件设计方案 9

2.2.2 系统软件设计方案 11

2.3 本章小结 12

第3章 车流量检测雷达数字信号处理系统的电路功能实现 14

3.1 数字信号处理系统总体结构设计 14

3.1.1 DSP核心处理器介绍及设计 14

3.1.2 系统总体结构设计 15

3.2 采样模块设计 17

3.2.1 采样电路设计 17

3.2.2 采样驱动软件设计 18

3.3 雷达调制信号产生模块的设计 21

3.3.1 调频三角波产生电路的设计 21

3.3.2 调频三角波驱动软件设计 22

3.4 通信模块设计 26

3.4.1 双通道串口电路设计 26

3.4.2 双通道串口通信驱动软件设计 30

3.5 外扩存储模块设计 33

3.5.1 程序存储模块 33

3.5.2 数据存储模块 36

3.6 本章小结 37

第4章 车流量检测雷达数字信号处理系统算法改进与实现 38

4.1 车流量检测算法中的目标信息提取和分析方法 39

4.2 提高信号采集模块性能的若干算法改进 41

4.3 提高目标判定模块性能的多次判定检测法 47

4.4 提高车流量检测系统回波信噪比的方法 50

4.5 本章小结 52

第5章 车流量检测雷达外场实验测试结果及系统改进方案 54

5.1 外场实验测试结果 54

5.2 系统改进方案 58

5.2.1 系统功能优化及扩展 58

5.2.2 基于TMS320VC5502的系统改进方案 64

5.3 本章小结 66

第6章 总结和展望 67

6.1 总结 67

6.2 展望 68

参考文献 69