基于多态数据采集的电力巡检驾驶舱故障预警与辅助决策

Cockpit fault early warning and assistant decision ofelectric patrol inspection based on polymorphic data acquisition

采用Pearson法、大数据挖掘法进行驾驶舱故障预警电力巡检时,存在电力巡检数据遗漏问题,导致故障预警准确率低,为此提出一种基于多态数据采集的电力巡检驾驶舱故障预警与辅助决策方法。设计驾驶舱故障预警与辅助决策结构图,采集电力巡检驾驶舱的多态故障数据;以多态数据为基础,判断电力系统故障的发生情况;通过数据预处理进行故障筛选,并以此为基础进行电力系统故障预警;根据EEAC理论生成驾驶舱故障辅助决策方法。为验证该方法的预警效果,设计了相关对比实验。结果表明,当迭代次数为2 000时,该方法的故障预警准确率为90%,故障预警效果较好。

When Pearson method and big data mining method are used to conduct power inspection for cockpit fault warning,there exists the problem of omission of power inspection data,which leads to low accuracy of fault warning.To this end,it proposes a method of early warning and auxiliary decision-making for power patrol cockpit based on polymorphic data collection,designs the cockpit failure early warning and auxiliary decision-making structure diagram,collects the polymorphic fault data of the electric inspection cockpit.Based on the polymorphic data,it judges the occurrence of faults in the operation of the electric inspection cockpit,realizes the cockpit fault screening,carries out the electric power inspection cockpit fault warning.According to the EEAC theory,the auxiliary decision-making control of the cockpit fault is generated.In order to verify the cockpit failure early warning effect,the relevant comparative experiments are designed.The experimental results show that when the number of iterations is 2000,the fault warning accuracy of this method is 90%,and the fault warning effect is good.