智能监测终端第三方测试数据
收藏资源简介：
本数据集主要面向港口长距离、高速带式输送系统智能运维技术的工程化应用研究与核心算法验证需求而建设。数据集源于无锡宝通科技股份有限公司智能实验室与武汉理工大学港机大车间两大实验平台。数据采集工作于2025年11月至12月期间进行，由具备CMA资质的第三方检测机构，采用“终端原生机+标准校验设备”的双设备采集模式，运用RFID读写器、红外热像仪、振动传感器等多种专业设备，模拟高速、长距离及各类故障工况，对9类智能监测终端及1类智能巡检机器人开展全面测试。本数据集主要记录了9类智能监测终端（包括数字化输送带身份识别、输送带纵撕监测、智能托辊监测、骨架无损监测、磨损测厚、滚筒包胶监测、智能纠偏、无线测温、定点停机终端）及1类智能巡检机器人的全流程第三方测试数据。数据内容涵盖结构化性能数据（故障识别率、报警响应时间、定位精度等）、物理量监测数据（温度、振动、速度、厚度等）以及非结构化影像数据（红外热像图、可见光故障截图、X射线骨架图像）。数据集经过严格的质量控制，包括缺失值插值补全、异常值剔除及标准化人工标注（如纵撕裂口、包胶异物位置标注），有效排除了设备误触发与环境干扰。数据集共包含1个核心数据包，内含10个子模块（对应各终端类型的.xlsx文件及图片文件夹），数据总量约248.94MB。该数据集真实反映了智能终端在复杂工况下的核心性能表现，填补了行业内缺乏统一第三方标准化测试数据的空白，可为带式输送系统从“定期检修”向“预测性维护”转型提供关键的数据验证与算法迭代支撑。
This dataset is developed to meet the requirements of engineering application research and core algorithm verification for intelligent operation and maintenance technology of long-distance, high-speed belt conveyor systems in ports. The dataset is sourced from two experimental platforms: the Intelligent Laboratory of Wuxi Baotong Technology Co., Ltd. and the Port Machinery Workshop of Wuhan University of Technology. Data collection was carried out from November to December 2025, conducted by a third-party testing institution with China Metrology Accreditation (CMA) qualification. Adopting the dual-device collection mode of 'terminal native device + standard calibration equipment', the test uses multiple professional devices including RFID readers/writers, infrared thermal imagers, vibration sensors, etc., to simulate high-speed, long-distance and various fault working conditions, and conduct comprehensive tests on 9 types of intelligent monitoring terminals and 1 type of intelligent inspection robot. This dataset mainly records the full-process third-party test data of the 9 types of intelligent monitoring terminals (including digital conveyor belt identification, conveyor belt longitudinal tear monitoring, intelligent idler monitoring, framework nondestructive monitoring, wear thickness measurement, drum rubber coating monitoring, intelligent deviation correction, wireless temperature measurement, fixed-point shutdown terminal) and 1 type of intelligent inspection robot. The data content covers structured performance data (such as fault recognition rate, alarm response time, positioning accuracy, etc.), physical quantity monitoring data (temperature, vibration, velocity, thickness, etc.), and unstructured image data (infrared thermal images, visible light fault screenshots, X-ray framework images). The dataset has undergone strict quality control, including missing value interpolation and completion, outlier removal, and standardized manual annotation (e.g., annotation of longitudinal tear openings and positions of foreign matters in rubber coating), effectively eliminating device false triggers and environmental interference. The dataset consists of 1 core data package, which contains 10 submodules (corresponding to .xlsx files and image folders for each terminal type), with a total data volume of approximately 248.94 MB. This dataset truly reflects the core performance of intelligent terminals under complex working conditions, fills the industry gap caused by the lack of unified third-party standardized test data, and can provide key data verification and algorithm iteration support for the transformation of belt conveyor systems from 'regular maintenance' to 'predictive maintenance'.
