高级检索

    在役螺栓超声三维成像监测数据的智能化分析方法

    Intelligent analysis method of ultrasonic 3D imaging monitoring data for bolts in service

    • 摘要: 螺栓是一种常见的机械连接件,一旦失效,可能会导致设备损坏、结构崩塌,甚至危及人身安全。因此,在役螺栓的监测一直受到广泛关注。将相控阵超声全聚焦技术应用于螺栓在役监测,开发了专用系统和阵列探头,实现了三维成像可视化监测,改变了螺栓监测数据单一的现状。针对大量三维图像化数据处理的关键问题,提出了智能化分析方法,对比分析了决策树、支持向量机和神经网络算法的适用性,最终开发了基于决策树算法的智能分析评价软件模块并取得了较好的检测效果。

       

      Abstract: Bolt is one of common mechanical connection parts. Once a bolt fails unexpectedly, it may cause equipment damage, structural collapse, or even endanger personnel safety. Therefore, the in-service monitoring of bolts has been widely concerned. In this paper, the phased array ultrasonic total focusing technology was applied to the in-service monitoring of bolts, and a special system and array probe were developed to realize the three-dimensional imaging visualization monitoring and change the current status of limited information content of the monitoring data. Aiming at the key issues of massive 3D image data processing, an intelligent analysis method was proposed, and the applicability of decision tree, support vector machine and neural network algorithms were compared and analyzed, an intelligent analysis and evaluation software module based on decision tree algorithm was finally developed and good detection results were obtained.

       

    /

    返回文章
    返回