塔机障碍物远距离超声测距方法与识别机理研究

发布时间：2018-01-24 12:45

本文关键词： 塔式起重机超声测距小波消噪回波幅值模型障碍物识别　出处：《西安建筑科技大学》2015年博士论文　论文类型：学位论文

【摘要】：塔式起重机(以下简称塔机)是建筑施工行业中的标志性装备,具有高空作业、重心高、工作幅度大的特点,广泛应用于建筑工地上的水平和垂直吊装。随着城市基础建设规模的不断扩大,塔机之间的碰撞以及塔机与周边障碍物的碰撞概率随之增高,由此造成的重大事故时有发生且呈上升趋势。针对目前塔机防碰撞监测技术存在的不足,本文基于主动式测障以及预警思想,研究塔机周边障碍物远距离超声传感测距以及障碍物超声回波特征识别方法,具有重要的理论和工程意义。本文完成的主要研究工作及创新性成果有:针对项目组研制的远距离超声测量硬件装置,构建了基于虚拟仪器的超声测量系统并对其进行了系统优化实验研究。设计了基于Lab VIEW的上位机软件,具有超声信号的采集与分析等功能;根据超声学和振动学理论,分析了超声信号回波过程及其产生机理;根据发射激励脉冲个数与超声回波幅值关系的实验,优化了远距离超声测量系统的发射参数。通过障碍物距离实验和超声波入射角实验,证明了该测量系统对30米范围内障碍物检测的有效性。针对传统数字滤波对峰值坐标有时间延迟且信噪比较低、影响测距准确度和目标信号的检测等问题,提出了适合分析非平稳超声信号的小波改进阈值消噪算法。根据超声回波信号噪声特点,确定了消噪评价指标。通过实验,分析了传统数字滤波算法在超声回波信号消噪前后的延时、信噪比和均方误差等方面的不足。根据小波阈值消噪的方法,提出了一种新的参数可调阈值函数。应用小波改进阈值消噪算法进行了仿真和实验研究,结果表明:改进阈值函数具有良好的消噪效果和灵活的调节性,提高了测量系统的抗干扰能力。研究了利用超声信号回波模型进行远距离测距的方法。利用模型仿真和实验装置详细分析了传统的阈值法和互相关法在远距离超声测距方面的抗噪性、精度、稳定性和适用性情况,针对二者测距存在的不足,研究了超声回波受迫上升段波形与超声激励脉冲的关系,建立了回波的包络曲线时域信号数学模型,并提出了改进的包络峰值法和峰值结合包络模型法两种测距方法。仿真和实验表明改进的测距算法的远距离测距性能优于传统的测距法,对单障碍物和多障碍物的远距离测距都能满足工程实际精度要求。研究了障碍物远距离超声回波幅值模型和识别机理。通过理论推导和实验分析,研究了传感器张角与距离的关系,得到不同距离的张角能量系数;研究了不同距离处对应的最大有效圆面积和障碍物面积的比例关系,得到面积比例系数;分析了角类、面类、柱类障碍物的回波特点,综合各影响因素,建立了典型材质和形状障碍物的超声回波幅值模型。分析了传感器入射角对回波的振动时间特征的影响,研究了传感器小倾角范围内不同材质引起的回波差异,对它们的回波进行时域和时频域特征提取,采用基于欧式距离判据的特征评价和支持向量机进行材质的分类识别,识别效果良好。
[Abstract]:Tower crane (hereinafter referred to as the tower crane) is the construction industry of the iconic equipment, with high-altitude operations, high center of gravity, large work range characteristics, widely used in the construction site on the horizontal and vertical lifting. With the continuous expansion of the scale of city infrastructure construction, the tower crane collision between the tower and the surrounding obstacles and the collision probability increased, resulting in major accidents have occurred and the upward trend. Aiming at the shortcomings of tower crane anti collision monitoring technology, the active measuring and warning based on the idea of the tower crane around the obstacle distance ultrasonic sensor ranging and obstacle ultrasonic echo feature recognition method has important theoretical and practical significance. In this paper, the main research work and innovative achievements are: according to the distance ultrasonic measurement hardware device developed by the research team, is constructed based on virtual instrument The ultrasonic measurement system and optimized experimental system. The design of the PC software Lab based on VIEW, data acquisition and analysis has the function of ultrasonic signal; according to the ultrasound and vibration theory, analyzes the process of ultrasonic echo signal and its mechanism; according to the emission excitation pulse number and amplitude of ultrasonic echo the emission parameters of ultrasonic measurement system was optimized. Through long-distance obstacle distance test and ultrasonic angle experiments demonstrate the validity of the measurement system for obstacle detection of a range of 30 meters. The traditional cataract digital filter with time delay and low SNR of peak coordinates, influence range accuracy and the target signal detection, wavelet analysis is proposed for non-stationary ultrasonic signal, the improved threshold de-noising algorithm. According to the ultrasonic echo signal noise characteristics, determine the noise evaluation index Through the experiment, analysis of the traditional digital filtering algorithm in the denoising of ultrasonic echo signal before and after the delay, lack of signal to noise ratio and mean square error and so on. According to the method of wavelet threshold denoising, a new threshold function with adjustable parameters. Application of improved wavelet threshold denoising algorithm is proved by simulation and experiment research results show that the improved threshold function has good denoising effect and flexible adjustment, improve the anti-interference ability of the measuring system. The research method of long distance measurement using ultrasonic echo signal model. Simulation and experimental device model based on detailed analysis of the traditional threshold method and anti noise, cross-correlation method in long range ultrasonic ranging in terms of accuracy, stability and applicability, to solve the problems existing in the two range, forced ascent waveform and ultrasonic pulse echo of the relationship, to establish the The envelope curve of time-domain signal wave mathematical model, and proposes an improved peak envelope method and the peak envelope binding model method two ranging method. Simulation and experimental results show that the improved algorithm of distance ranging distance and better performance than the conventional, far distance of single obstacle and multiple obstacles can meet the requirements of the practical accuracy Study on the obstacle distance. The ultrasonic echo amplitude model and recognition mechanism. Through theoretical derivation and experimental analysis, studied the relationship between the angle and distance of the sensor, get the energy coefficient of different distance angle; studied at different distances corresponding to the maximum effective circle area and obstacle area ratio, get the area proportion coefficient; analysis angle, plane, echo characteristics of column class barriers, the influence factors, established a typical material and shape of the obstacle ultrasonic echo amplitude model. Analysis of the influence of vibration sensor time characteristics of incident angle on the echo, echo of different materials caused by sensor of small angle range, the echo of the time domain and frequency domain feature extraction, classification and identification of the material characteristic evaluation criterion of Euclidean distance and support vector machine based on the identification results.

【学位授予单位】：西安建筑科技大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：TB559

【相似文献】