六面顶压机顶锤破裂的在线无损检测研究

发布时间：2018-06-02 13:05

本文选题：六面顶压机 + 顶锤破裂　；参考：《郑州大学》2015年硕士论文

【摘要】：六面顶压机顶锤破裂是人造金刚石合成过程中常见的一种生产事故,若未及时发现,极易造成更大的经济损失,甚至人员伤亡。但当前国内外并无顶锤破裂在线检测的相关应用产品,现阶段的检测手段主要是操作工手工划锤及有经验的工人在生产过程中听音识别。因此顶锤破裂的在线检测具有很强的现实意义,同时也是人造金刚石合成的工业自动化发展中的一个关键问题。本文通过对顶锤破裂成因的分析,结合顶锤工作环境,提出了一种基于声发射的顶锤破裂在线检测方案。结合当前人工听音识别方法,将声发射检测信号确定在音频段,并从理论上论述其合理性,利用工控机声卡完成信号采集。完成声发射检测模块的设计,包括压电换能元件的选取、外壳设计、电源模块设计以及信号前置放大和滤波电路等信号调理硬件电路设计。利用该传感检测模块采集现场顶锤工作信号,直观分析可以将其分为3类:低能量的噪音、疑似信号(顶锤预紧环由于加热温度波动而热胀冷缩及合成压力波动所引起的震动信号)、裂锤信号。对顶锤破裂信号特征进行前期研究。采用频谱分析、STFT及EMD经验模态分解方法分析,发现疑似信号在200Hz~400Hz、2KHz~4KHz频率段特征明显,采用EMD经验模态分解取得的HHT边际谱清晰显示裂锤故障频率存在于13KHz~16KHz。通过C++和Matlab混合编程方法完成系统软件设计。采用C++编写软件程序,包含声卡信号采集、图形界面显示及多线程信号分析框架等。设计基于短时能量和短时平均过零率双门限的端点检测算法,截取出疑似信号和裂锤信号,完成信号预处理。信号识别算法利用Matlab语言编写,编译成COM组件,在信号分析线程调用完成裂锤信号识别。本文设计的检测系统目前继续进行现场信号采集,同时利用现有样本特征,设计信号识别算法,在现场进行测试,当出现误报警或破裂未报警时,提取信号进行分析,改进信号识别算法。
[Abstract]:The breakage of the top hammer of a six-sided press is a common production accident in the synthetic process of synthetic diamond. If it is not discovered in time, it is easy to cause greater economic losses and even casualties. However, there are no related products for on-line detection of top hammer rupture at home and abroad. At present, the main means of detection are manual stroke hammer and experienced workers in the production process. Therefore, the on-line detection of hammer rupture is of great practical significance and is also a key problem in the development of industrial automation of synthetic diamond synthesis. Based on the analysis of the causes of top hammer rupture and combined with the working environment of top hammer, an on-line detection scheme of top hammer rupture based on acoustic emission is proposed in this paper. Combined with the current method of artificial audio recognition, the acoustic emission detection signal is determined in the audio section, and its rationality is discussed theoretically, and the signal acquisition is completed by using the sound card of the industrial control computer. Complete the design of acoustic emission detection module, including the selection of piezoelectric energy transfer elements, shell design, power module design and signal preamplifier and filter circuit signal conditioning hardware circuit design. The sensor module is used to collect the working signals of the top hammer, which can be divided into three categories: low energy noise, low energy noise, Suspected signal (top hammer pretightening ring due to heating temperature fluctuation caused by thermal expansion and contraction and synthetic pressure fluctuation caused by vibration signal, crack hammer signal. The characteristics of the breaking signal of the top hammer are studied in the early stage. By using spectrum analysis and EMD empirical mode decomposition method, it is found that the characteristic of suspected signal is obvious in the frequency range of 200Hz / 400Hz / 2kHz / 4kHz. The marginal spectrum of HHT obtained by EMD empirical mode decomposition clearly shows that the fault frequency of crack hammer exists at 13kHz / 16kHz. The system software is designed by C and Matlab mixed programming method. C is used to write software program, including sound card signal acquisition, graphical interface display and multi-thread signal analysis framework. An endpoint detection algorithm based on the double threshold of short time energy and short time average zero crossing rate is designed to intercept the suspected signal and crack hammer signal and complete the signal preprocessing. The signal recognition algorithm is compiled by Matlab language and compiled into a COM component, which is called by the signal analysis thread to complete the signal recognition of the hammer. The detection system designed in this paper continues to collect the field signals at present. At the same time, using the existing sample features, the signal recognition algorithm is designed and tested on the spot. When there is false alarm or failure to alarm, the signal is extracted for analysis. Improved signal recognition algorithm.
【学位授予单位】：郑州大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TQ163;TG115.28

【相似文献】