田湾核电站基于边缘计算与大数据的设备状态监测与智能诊断平台的应用

Application of Equipment Condition Monitoring and Intelligent Diagnosis Platform Based on Edge Computing and Big Data in TNPS

针对传统的设备状态监测主要依靠人工定期采集设备的振动参数和油质参数来实现有限的状态监测,存在功能不完善且准确率低,无法准确预测设备的异常和健康状态的问题,田湾核电站以上充泵、风机等设备为依托对象,探索建立设备监测与预测性诊断平台。平台借助无线传感技术、物联网技术、大数据分析技术,融合设备已有测点参数以及新增的振动、声发射、油液、电流及压力等传感器测点高频数据,设计由传感器、边缘计算网关、数据存储层、数据模型层、应用功能层、展示层组成的平台架构,通过频域分析算法和时域分析算法建立参数监测预警模型,实施设备状态监测、数据分析、性能预测、提前预警、智能诊断以及自动报告等功能,有效提升设备管理和故障诊断的准确性和效率,发挥了较好作用。

Aiming at the problem that traditional equipment condition monitoring mainly relies on manual periodic collection of equipment vibration parameters and oil quality parameters to achieve limited condition monitoring, which has imperfect functions and low accuracy, and can not accurately predict the abnormal and healthy state of equipment, Tianwan Nuclear Power Station’s above-mentioned pumps, fans and other equipment are supported by exploring the establishment of equipment monitoring and predictive diagnosis. The platform uses wireless sensing technology, Internet of Things technology, and big data analysis technology to integrate the existing measurement point parameters of the equipment and the new high-frequency data of the vibration, acoustic emission, oil, current, and pressure sensor measurement points. The platform architecture is composed of sensors, edge computing gateway, data storage layer, data model layer, application function layer and display layer. The parameter monitoring and early warning model is established by frequency domain analysis algorithm and time domain analysis algorithm.The functions of equipment condition monitoring, data analysis, performance prediction, early warning, intelligent diagnosis and automatic report are implemented. It effectively improve the accuracy and efficiency of equipment management and fault diagnosis.