| 287 | 3 | 16 |
| 下载次数 | 被引频次 | 阅读次数 |
本文通过STM32完成超声波的发射,回波接收及处理,然后对回波信号进行小波预处理,得到信号的幅频特性并分离出其高频分量和低频分量,以高频和低频分量作为信号特征,运用机器学习的Linear SVM算法对信号进行分类,以区分出前方物体是杂草还是树木,达到超声波智能识别的目的.
Abstract：In this paper,ultrasonic transmit,echo reception and process are all completed by STM32. Then the echo signal is disposed by wavelet pretreatment to get the signal amplitude frequency characteristics and isolate the high frequency and low frequency components. With high frequency and low frequency components as signal characteristics,we can use the Linear SVM algorithm of machine learning to classify signals so that to distinguish whether the front object is weeds or trees. As a result,the ultrasonic intelligent identification is achieved.
[1]赵连睿.超声相控阵自动测距系统的设计[D].济南:山东师范大学,2013.
[2]李建平,等.小波分析的一些有前景的应用领域[J].重庆大学学报:自然科学版,1999,(01):123-127.
[3]Majkowski A M,Koodziej,Rak R J.Joint time-frequency and wavelet analysis-an Introduction[J].Metrology and Measurement Systems,2014.21(4):741-758.
[4]田立,刘振丙,刘小茂,一种改进的线性SVM[J].计算机工程与应用,2007,(20):173-176.
[5]曹军,仪器放大器AD620性能及其应用[J].电子器件,1997,(03):64-68.
[6]刘甜,基于CFOA的快速响应滤波器的设计与实现[D].济南:山东师范大学,2016.
[7]闻明,吉海彦.小波分析消噪及其在Matlab中的实现[J].现代电子技术,2003,(24):47-49.
[8]张静远,张冰,蒋兴舟.基于小波变换的特征提取方法分析[J].信号处理,2000,(02):156-162+155.
[9]Iino Y,et al.,Model based extended wavelet analysis and its application to fault diagnosis of an elevator system[J].IFAC Proceedings Volumes,1997.30(11):621-626.
[10]溪海燕,肖志涛,张芳.基于线性SVM的车辆前方行人检测方法[J].天津工业大学学报,2012,(01):69-73.
基本信息:
DOI：
中图分类号:TP181;TU986.32
引用信息:
[1]侯东伟,杨济民,韩胜亮.用于庭院除草机的超声波智能避障系统的设计[J].山东师范大学学报(自然科学版),2018,33(01):57-60+68.
基金信息:
山东省重点研发计划项目(GG201703150121)