黄中生,李靖祥,赵升吨.基于振动信号的汽车纵梁腹面冲漏孔智能在线检测技术[J].精密成形工程,2022,14(7):28-35.
HUANG Zhong-sheng,LI Jing-xiang,ZHAO Sheng-dun.Intelligent Online Detection Technology for Missing Punching of Machining Holes on Ventral Surface of Automobile Carling Based on Vibration Signal[J].Journal of Netshape Forming Engineering,2022,14(7):28-35. |
| 基于振动信号的汽车纵梁腹面冲漏孔智能在线检测技术 |
| Intelligent Online Detection Technology for Missing Punching of Machining Holes on Ventral Surface of Automobile Carling Based on Vibration Signal |
| |
| DOI：10.3969/j.issn.1674-6457.2022.07.004 |
| 中文关键词: 汽车纵梁 漏孔 智能检测 振动信号 |
| 英文关键词: automobile carling missing hole intelligent detection vibration signal |
| 基金项目: |
|
| 摘要点击次数: 1602 |
| 全文下载次数: 858 |
| 中文摘要: |
| 目的 针对传统汽车纵梁腹面冲孔过程中产生的加工孔漏冲现象，提出一种基于振动信号的智能漏孔在线检测技术。方法 在纵梁冲孔设备上合理布置加速度传感器，对采集的振动信号进行时域和频域特性分析，构建归一化的压缩混合域特性指标矩阵，提取正常冲孔和漏冲信号的可靠评价指标，并进一步建立漏孔的智能在线检测方案。结果 在时域上，漏冲信号的平均值和有效值要比正常冲信号的低，同时具有明显冲击峰的波峰个数要少;在频域上，漏冲信号的频率幅值最大值、频率幅值平均值以及能量要比正常冲信号的低，但是其变异系数要比正常冲信号的高。结论 提出了一种智能漏孔检测方案，可以通过对采集的信号进行预判断、处理、特征提取、检测判断等，分离出漏冲信号和正常冲信号。 |
| 英文摘要: |
| The work aims to propose an intelligent online leak detection technology based on vibration signal to solve the phenomenon of missing punching of machining holes in the traditional automobile carling ventral punching process. By reasonably arranging the acceleration sensors on the carling punching equipment, the time domain and frequency domain characteristics of the collected vibration signals were analyzed, a normalized compressed mixed domain characteristic index matrix was constructed, and the reliable evaluation indicators for normal punching and missing punching signals were extracted, and an intelligent online detection scheme for missing punching was further established. Through analysis of the vibration signal in this project, it can be judged that in the time domain, the average value and effective value of the missing punching signal were lower than those of the normal punching signal, and the number of peaks with obvious impulse peaks was less, in the frequency domain, the maximum frequency amplitude, the average frequency amplitude and the energy of the missing punching signal were lower than those of the normal punching signal, but the coefficient of variation was higher than that of the normal punching signal. According to the experimental results, this work proposes an intelligent missing punching detection scheme, which separates the missing punching signal and the normal punching signal by prejudging, processing, feature extraction, detection and judgment on the collected signals. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
渝公网安备 50010702501719号