基于边缘计算与时序预测的海上平台电气安全监测预警系统研制

Development of Electrical Safety Monitoring and Early Warning System for Offshore Platform Based on Edge Computing and Time Sequence Prediction

为解决目前我国海上平台电气安全管控过程中缺乏低压交流系统中早期隐患预警能力的问题,研制了海上平台低压交流系统电气安全监测预警系统。该系统采用边云协同架构,在各级配电箱设置具有中早期隐患捕捉能力的边缘计算型监测终端,并在云端部署具有时序预测算法的隐患预警平台。该系统针对复杂电磁环境、耐盐性能和可靠性进行了加强设计,以适应海上平台场景潮湿、高盐和人力资源稀缺的环境特点。经测试试点,该系统报警准确率高于99.95%,投入使用后电气安全故障溯源排查平均时间从20h下降至30min,所有试点项目渐变性因素引发的电气安全事故数为0,达到了管控电气安全中早期隐患的效果,同时提升了电气维护人员的工作效率。

In order to solve the problem of lack of early warning ability of low-voltage AC system in the process of electrical safety control of offshore platforms in China,an electrical safety monitoring and early warning system of low-voltage AC system of offshore platforms is developed.The system adopts a side-cloud collaborative architecture,the edge computing monitoring terminal with the ability of catching the early and middle hidden dangers is set up in the distribution boxes at all levels,and the hidden danger warning platform with the time series prediction algorithm is deployed in the cloud.The system is designed for the complex electromagnetic environment,salt resistance and reliability to adapt to the environment characteristics of humid,high salt and scarce human resources in offshore platform scene.After testing,the alarm accuracy of the system is higher than 99.95%.After putting into use,the average time of electrical safety fault tracing and troubleshooting is reduced from 20 hours to 30 minutes.The number of electrical safety accidents caused by gradual factors in all pilot projects is 0,which achieves the effect of controlling early hidden dangers in electrical safety and improves the working efficiency of electrical maintenance personnel.