基于边缘计算的电子井盖异常检测技术优化

Optimization of electronic manhole cover anomaly detection technology based on edge computing

针对电子井盖的异常数据检测问题,研究提出了基于边缘计算的处理方式,提高电子井盖系统的运算效率。对边缘计算的数据处理方法进行优化,利用时空相关性相结合的方式,大大提高边缘计算的异常数据检测效率。实验结果表明,将所提算法与其他算法进行对比,该算法在1000个传感器的数据上,本文算法的运行时间约为0.25 s,远远优于其他算法,在电池能耗、数据传输方面同样优势明显。在井盖位移和水位测试等方面,电子井盖的误报率低于2%,具有重要的应用价值。

In order to solve the problem of abnormal data detection of electronic manhole cover,a processing method based on edge computing was proposed to improve the computing efficiency of the electronic manhole cover system.The data processing method of edge computing was optimized,and the method of combining spatio⁃temporal correlation was used to greatly improve the efficiency of anomaly data detection of edge computing.The experimental results showed that the algorithm had a running time of approximately 0.25 seconds on data from 1000 sensors,which was much better than other algorithms.It also had significant advantages in battery energy consumption and data transmission.In terms of displacement and water level testing of manhole covers,the false alarm rate of electronic manhole covers was less than 2%,which had important application value.