Scholarlink

<p>The hydraulic fine blanking press (FBP) is an advanced piece of manufacturing equipment whose operational state variations can impact the precision and quality of components, potentially causing die damage and hydraulic system failures. Consequently, the development of an intelligent and precise monitoring approach is crucial for this high-performance manufacturing equipment. In this study, we propose an intelligent monitoring methodology for the operational state of a hydraulic FBP based on fiber Bragg grating (FBG) sensors. Taking the HFB-500 hydraulic FBP as the subject of study, FBG sensors measured the body strain under different process parameters and continuous strain data from the slider. The strain data was denoised using db4 wavelet soft thresholding in MATLAB. Both time-domain and frequency-domain analyses were conducted to examine the characteristics of the strain data under different process parameters. A Bayesian-optimized Random Forest (BO-RF) state classification model was developed based on time-domain features and compared with Support Vector Machine (SVM) and K-Nearest Neighbors (KNN) models. The analysis of the experimental data indicated that when the blank holder force and counter punch force were set to 600 kN and 200 kN, exhibited minimal time-domain feature dispersion and concentrated peak frequencies. The proposed method achieved a classification accuracy of 99.27%. The analysis of the slider data indicated that three types of random problems were successfully identified: excessive scrap on the die surface, feeder jamming, and parts getting stuck in the middle of the die, proving that the method works well for accurate monitoring in manufacturing settings.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

State monitoring of hydraulic fine blanking press based on FBG sensors

  • 期刊论文

  • 发表日期：

    2026年4月2日

查看全文
  • Wenting Wei,
  • Shuzhao Zhang,
  • Yanxiong Liu,
  • Lin Hua

摘要

The hydraulic fine blanking press (FBP) is an advanced piece of manufacturing equipment whose operational state variations can impact the precision and quality of components, potentially causing die damage and hydraulic system failures. Consequently, the development of an intelligent and precise monitoring approach is crucial for this high-performance manufacturing equipment. In this study, we propose an intelligent monitoring methodology for the operational state of a hydraulic FBP based on fiber Bragg grating (FBG) sensors. Taking the HFB-500 hydraulic FBP as the subject of study, FBG sensors measured the body strain under different process parameters and continuous strain data from the slider. The strain data was denoised using db4 wavelet soft thresholding in MATLAB. Both time-domain and frequency-domain analyses were conducted to examine the characteristics of the strain data under different process parameters. A Bayesian-optimized Random Forest (BO-RF) state classification model was developed based on time-domain features and compared with Support Vector Machine (SVM) and K-Nearest Neighbors (KNN) models. The analysis of the experimental data indicated that when the blank holder force and counter punch force were set to 600 kN and 200 kN, exhibited minimal time-domain feature dispersion and concentrated peak frequencies. The proposed method achieved a classification accuracy of 99.27%. The analysis of the slider data indicated that three types of random problems were successfully identified: excessive scrap on the die surface, feeder jamming, and parts getting stuck in the middle of the die, proving that the method works well for accurate monitoring in manufacturing settings.