面向贮箱超大结构立式装配对接面几何特征的高精度在机测量数据集

本研究构建了一个面向焊接过程中的在机测量数据集，旨在为焊接对接面几何特征的高精度获取与测量控制提供数据支持。数据集涵盖了高温环境下的测量误差修正、线激光光条图像去模糊、高反光表面光条中心提取和焊接现场对接面几何特征测量等典型应用场景。数据采集基于多平台实验，涉及电磁感应加热控制的高温表面、振动实验模拟的模糊光条图像、高反光表面曝光控制以及火箭贮箱筒段的现场焊接测量。所有数据经过严格的清洗、标注和异常处理，确保数据的准确性和完整性，并通过标准化处理、光谱共焦传感器验证和第三方机构检测，确保数据质量。该数据集在焊接过程的几何误差检测、质量预测及工艺优化中具有广泛应用价值，能够为焊接过程的智能化调控与优化决策提供高质量数据支撑，推动焊接技术从经验驱动向数据驱动的精准化、智能化转型。This study constructs an in-process measurement dataset for welding processes, aiming to provide data support for high-precision acquisition and measurement control of geometric features of welded butt joints. The dataset covers typical application scenarios including measurement error correction in high-temperature environments, deblurring of line laser stripe images, extraction of stripe centers on highly reflective surfaces, and geometric feature measurement of on-site welded butt joints. Data collection is based on multi-platform experiments, involving high-temperature surfaces controlled by electromagnetic induction heating, blurred stripe images simulated via vibration experiments, exposure control on highly reflective surfaces, and on-site welding measurement of rocket tank barrel sections. All data has undergone strict cleaning, annotation and outlier processing to ensure the accuracy and integrity of the dataset. The data quality is guaranteed via standardization processing, verification with spectral confocal sensors and inspection by third-party institutions. This dataset has broad application value in geometric error detection, quality prediction and process optimization during welding processes. It can provide high-quality data support for intelligent regulation and optimal decision-making of welding processes, and promote the transformation of welding technology from experience-driven to data-driven precision and intelligent upgrading.

搜集汇总

背景与挑战

背景概述

该数据集旨在支持焊接对接面几何特征的高精度测量，涵盖高温环境下的误差修正、光条图像去模糊等典型应用场景。数据采集涉及多平台实验，并经过严格清洗和验证，确保数据质量，为焊接过程的智能化调控提供支撑。