• CN:11-2187/TH
  • ISSN:0577-6686

›› 2013, Vol. 49 ›› Issue (18): 184-189.

• 论文 • 上一篇    下一篇



  1. 江苏大学电气信息工程学院;江苏省(扬州)数控机床研究院
  • 发布日期:2013-09-20

High Pressure Water Jet Cutting Sound Feature Extraction and Nozzle Traverse Speed Control

ZHANG Jun;ZHAO Dean;LI Xuefeng;XU Lilong   

  1. Electrical and Information Engineering College, Jiangsu University CNC Machine Tool Research Institute of Jiangsu Province
  • Published:2013-09-20

摘要: 超高压水射流喷嘴移动速度是影响水射流切割质量和效率的关键因素,如何对不同的工件选择最优的喷嘴移动速度是实现水射流柔性切割的主要研究内容。当超高压水射流切割不同材质和几何形状的目标物体时,水射流反射声音特征量也不同,包括反射声音能量和切割声音持续长度,工件越厚,切割时间越长,声音能量越大,反射声音特征量间接反映了工件切割的难易程度,这为优化喷嘴移动速度提供了基础。设计水射流切割声音计算机采集系统,采集的切割反射声音包括背景噪声和工件切割反射声音,基于第二代小波变换和Teager能量算子对水射流切割的反射声音信号处理,抑制背景噪声的影响,提取反映切割不同工件质量的声音特征;利用最小二乘向量机方法对切割喷嘴移动速度和声音特征量进行非线性建模,实现水射流切割的自适应速度优化控制,减少人工经验的试凑,提高生产效率。通过现场多次试验验证所提试验方案的有效性。

关键词: 超高压水射流, 喷嘴移动速度, 声音特征, 小波分析, 支持向量机

Abstract: The key factor to influence the quality and efficiency of water jet cutting is the traverse speed of ultra-high pressure water jet nozzle. The major research to achieve flexible water jet cutting is about how to choose the optimal nozzle traverse speed to different workpieces. When the ultra-high pressure water jet cuts different geometric objects made of various materials, the reflected sound feature are different, which mainly consist of the reflected sound energy and cutting sound time. As the workpieces thickness increases, both the reflected sound energy and cutting time will raise at the same time. Indirectly, complexity and difficulty of workpieces cutting are indicated by the reflected sound feature, which provides convenience for optimizing nozzle traverse speed. In this paper, a computer operated acquisition system for water jet cutting sound is designed. In which the reflected cutting sound signal, mainly including background noises and the reflected sound signal of workpieces cutting, is based on that the second generation wavelet transform and Teager energy operator are applied organically conjunctively in processing cutting sound signal for denoising background noises, and to extract sound features which reflect the quality of cutting different workpieces. Moreover, the nonlinear model is established between the cutting nozzle traverse speed and the sound feature based on least squares support vector machine, reduce artificial experience of trial and error, and enhance production efficiency. In several field tests, the feasibility and validity of the proposed scheme are verified and confirmed.

Key words: High pressure water jet, Nozzle traverse speed, Sound feature, Support vector machine, Wavelet analysis