【作者】 郎需强；

【导师】 侯加林；

【作者基本信息】 山东农业大学 ， 农业机械化工程， 2014， 博士

【摘要】 昆虫运动机理仿生研究有望为微型仿生机器人的设计提供新的思路和理论指导，因而自然生活条件下昆虫运动学参数的测量对该项研究有着至关重要的意义。在昆虫运动参数的测量研究中，昆虫的实际生活环境和实验室搭建的探测环境之间的差异成为该领域进一步发展的瓶颈。提出一种新的昆虫运动特性获取方法是解决这一瓶颈的根本之策。本论文提出了基于微多普勒效应的昆虫运动特性获取系统，利用该系统不但可以对自然生活条件下的昆虫运动形态进行探测，还能对昆虫躯体各个组成部分的运动行为进行细致描绘。本文首先从昆虫运动学信息的获取方法阐述了在昆虫运动学信息获取方面存在的问题。然后结合雷达在昆虫学领域的应用，以及微动多普勒雷达在空间、地面、生物体运动特征以及人体生命特征中的发展现状，提出从昆虫运动雷达多普勒回波中，获取其运动特性所需要研究的问题。并对点目标模型微多普勒回波方程以及时频分析和特征提取方法、线形昆虫躯体运动对雷达回波的调制作用。昆虫微多普勒特征分析和运动模式的识别等问题进行探索性的研究。针对点目标模型微多普勒回波方程以及时频分析和特征提取方法问题。首先在基于电磁波回波理论和目标散射特性的基础上，利用正交检波器和带通滤波器获取雷达回波中的微多普勒信号。然后利用点目标微动多普勒回波一般形式，推衍出点目标间歇振动的雷达回波解析表达式，并通过一阶贝塞尔函数对其回波形式进行分析。随后利用STFT、WVD、SPWVD、RSPWVD时频分析方法，对多个模型点回波仿真信号进行时频分析。通过时频谱图的比较体现本文选用的RSPWV分析方法的优越性。对时频谱图的特征提取进行了阐述。随后将经过RSPWV变换后的多个散射点回波时频谱图，分别用一阶时间条件矩、峰值检测、基于Lablacian图像脊线法对其特征进行提取和比对。验证基于Lablacian图像脊线法对包含噪声的多探测目标点回波时频谱图进行特征提取的优点。针对线形昆虫在运动过程中对电磁波的调制作用问题。首先在基于类蛇形机器人“铰链”结构和系统所采用的雷达工作方式的基础上，提出了适用于线形昆虫雷达特征研究的结构模型。并将以黄粉虫为代表的线形昆虫躯体结构分为头部、腹部、尾部三个部分。通过分析获取黄粉虫躯体不同部位的运动形式，结合点目标雷达回波方程推衍出线形昆虫躯体不同部位运动的雷达回波解析表达式。然后运用时频分析工具和特征提取方法，获取黄粉虫不同部位雷达回波中的微多普勒特征，进而对黄粉虫各个躯体部分时频分量进行研究。为支撑本论文在黄粉虫运动多普勒回波仿真的研究工作，设计并完成了工作在K波段的运动目标体微多普勒雷达探测系统。首先对运动目标体微多普勒雷达探测系统的需求进行了剖析，提出了雷达系统的组成结构。随后完成了雷达探测系统的硬件设计，着重对微多普勒雷达射频前端（微带天线仿真和设计、系统接收和发射端的增益控制以及多普勒回波信号的提取和进一步处理）的设计进行了阐述。然后在获取到雷达多普勒回波信号后，由基于STM32的中央处理单元实现回波信号的实时采集、储存和与上位机的数据通信。最后在上位机端利用MATLAB中的时频分析工具箱实现回波信号的时频分析。针对黄粉虫微多普勒特征分析和运动模式识别的问题，本文利用搭建的运动目标体微多普勒雷达探测平台，对不同运动形态下的黄粉虫进行微多普勒信号回波采集。为达到良好的电磁屏蔽效果，本文利用吸波材料搭建的微波暗室进行回波信号的屏蔽。然后利用多普勒雷达探测系统对单个黄粉虫头部摆动、尾部摆动、缓慢爬行等运动形式和两个、多个黄粉虫运动多普勒回波进行探测。并完成回波信号的时频分析和特征提取，从中获取包含黄粉虫躯体各运动部分的回波特征，支持本课题的研究。最后对该雷达探测平台在探测过程中暴露的一些问题进行分析，并提出了下一步的改进方案。最后，对该课题主要研究的工作进行总结，并对该课题研究内容的发展趋势、应用前景进行了分析，指出需要进一步研究和解决的问题。本文主要工作目的是提出一种新型的昆虫运动学信息获取新方法，并讨论了利用该种方法进行昆虫运动学信息获取的可行性。由于该方法涉及的领域比较广，且都为高频电磁波前言技术，因此本文工作只是探索性的，在一定程度上只能算是初步的阶段性工作。今后在对不同类型昆虫运动学信息获取系统时，还有不少的问题有待于进行进一步解决和更深入的研究。更多还原

【Abstract】 As the bionic study of insect movement mechanism is expected to provide new ideas andtheoretical directions for the design of micro bio-robot, the measurement of insect kinematicsparameter in natural condition has a magnificent significance for this study. In the study of themeasurement of insect kinematics parameter, the difference between insect’s livingsurroundings and the detect environment of laboratory building has became the bottleneck forthe further development, so a new way to acquire insect kinetic characteristics is thefundamental solution. This paper proposes insect kinetic characteristics capturing systembased on micro Doppler principle, which can not only detects insect kinetic characteristics,but also depicts the motor behaviors of all parts of insect’s body in detail.First, this paper elaborates the existing problems in the information acquisition of insectkinematics from the point of its acquiring method. Then, in accordance with the application ofradar, as well as the development situation of micro Doppler Radar in space, ground,organism’s motion feature and human’s vital signs, this paper proposes the relevant issues ofacquiring insect kinetic characteristic in echoes of Doppler Radar. What’s more, it also coversthe objective model of micro Doppler echo equation, time-frequency analysis and extractionmethod of characteristics as well as the modulating actions of radar echo for the linear insectbody movement. Finally, the exploratory study on insect’s micro Doppler feature analysis andthe recognition of motor pattern.In allusion to the objective model of micro Doppler echo equation and time-frequencyanalysis and extraction method of characteristics, first, we can take advantage of quadraturedetector and band-pass filter to acquire the micro Doppler signals based on electromagneticwave echo theory and scattering properties of the targets. Then, by virtue of the general typeof micro Doppler echo for the point target, we can deduce its analytical expression created byblocking vibration and analyze the waveform through the first order Bessel function. Takingadvantage of time-frequency analysis methods, such as STFT、WVD、SPWVD、RSPWVD,many point target echo simulations can be analyzed. Reflect the superiority of choosingRSPWV analytical method through the comparison of time-frequency spectrum. Finally,elaborate the characteristics of time-frequency spectrum, compare its characteristics throughfirst order time-conditional moment, peak detection and Image baseline method of Lablacian,and verify the merits extracted from the echoing spectrogram including noise based on Image baseline method of Lablacian.As for the modulating action of electromagnetic wave for the linear insect bodymovement, first, on the basis of the “hinge” construction of snake-shaped robot and workingmode of radar utilized by the system, we propose structural model that are suitable for thestudy of linear insect’s radar characteristics. Taking Tenebrio molitor as an example, linearinsect’s body can be divided into head, abdomen and tail. Then, through analyzing the motionform of different parts of Tenebrio molitor’s body, as well as its radar echo analyticalexpression deduced from the point target radar echo equation, we can acquire micro Dopplercharacteristics in the radar echoes of different parts of Tenebrio molitor and do the study on itstime-frequency component by utilizing time-frequency analytical tools and feature extractionmethod.To support the research work in this paper of the Doppler echo simulation of Tenebriomolitor motion, the micro Doppler radar detection system is designed and completed ofmoving targets operating at K band. Firstly, it conducts the demand analysis of the microDoppler radar system of moving targets, and puts forward the composition and structure ofradar system. Then it completes the hardware design of radar detection system, focusing onthe design of micro Doppler radar radio frequency front end (the simulation and design ofmicrostrip antenna, system receipt and gain control of transmitting terminal, and extractionand further processing of Doppler echo signal) to set forth here. And then after the access tothe radar Doppler echo signal, it realizes the real-time acquisition, storage, and datacommunication of host computers of echo signal based on the central processing unit ofSTM32. Finally, it realizes time-frequency analysis of the echo signal by the use of thetime-frequency analysis toolbox of MATLAB in host computers.Aiming at the micro Doppler feature analysis and pattern recognition problems of theTenebrio molitor, this paper, by use of structures of the micro Doppler radar detectionplatform for the moving target, collects the micro Doppler echo signal of the Tenebrio molitorunder different forms of motion morphology. In order to achieve a good electromagneticshielding effect, this paper uses the anechoic chamber constructed by absorbing materials toshield the echo signal. Then it detects by advantage of the Doppler radar detection system themotion forms of the head swinging, tail swinging and crawling movement of single Tenebriomolitor and the Doppler echo of the movement of two or many Tenebrio molitors. And itcompletes the time-frequency analysis and feature extraction of echo signal, in which gets theecho characteristics of each movement parts of the Tenebrio molitor body, to support theresearch of this task. Finally, it will analyze some problems exposed in the detection process of the detection platform with the radar, and put forward improving plan for the next step.Finally, it will be the summary of the main research work, and analyze the developmenttrend of the research content and application prospect, thereby it will point out the problemsthat need further study and solve.The main work of this paper is to propose a new method of the information acquisitionmodel of insect kinematics, and to discuss the feasibility of the information acquisition ofinsect kinematics by using the method. Because this method involves a relatively wide field,and are all the frontier technologies of the high-frequency electromagnetic wave, the work ofthis paper is just exploratory, to a certain extent, which can only be regarded as the initialstage. In the future, for different types of insect kinematics information acquisition system,there will be many problems to be studied and further resolved.更多还原