基于动态阈值的稳态模拟量异常预警方法

A Method for Abnormal Warning of Steady-state Analog Based on Dynamic Thresholds

在对生产生活中常见的电压、电流、温度等稳态模拟量进行异常预警分析时,当前监测类系统存在阈值维护工作量大、算法模型技术复杂和因单点异常判定导致了频繁预警和漏预警等问题。针对上述问题,本文提出了一种基于动态阈值和滑动窗口式分析的稳态模拟量异常预警方法。该方法能提高监测类系统预警功能的易维护性和预警有效性。

For warning analysis of steady-state analog data such as voltage,current,and temperature,which are common in production and life,the current monitoring system have problems of large threshold maintenance workload,complex algorithm model technology,and frequent or missed warnings caused by single point anomaly recognition.In order to overcome these problems,this paper proposes a method for steady-state analog abnormal warning based on dynamic threshold and sliding window analysis,which could enhance the maintainability and effectiveness of the warning function in the monitoring system.