摘要
针对传统巡检线路故障诊断方法中存在诊断准确率较低的问题,提出基于动态贝叶斯网络的机器人巡检线路故障诊断方法。利用巡检机器人的传感装置进行线路基础数据的采集,并进行融合处理。根据融合结果进行机器人线路运行状态检测,同时搭建动态贝叶斯网络故障诊断模型。利用诊断模型对机器人线路运行状态异常的线路,从输电线路故障和机器人运行故障两个方面,进行巡检线路故障类型的确定,最后对巡检线路故障发生概率进行计算,实现机器人巡检线路故障诊断结果的量化。通过对比实验的分析得出:与传统方法相比,所提方法的平均诊断准确率为93.4%,提升了10.8%。
Aiming at the problem of low diagnostic accuracy in traditional circuit fault diagnosis methods,a robot circuit fault diagnosis method based on dynamic bayesian network is proposed.The sensor device of inspection robot is used to collect the basic data of the line,and the data is fused.According to the fusion results,the robot circuit running state is detected and the dynamic bayesian network fault diagnosis model is built.The diagnosis model is used to determine the fault types of circuit inspection from the two aspects of transmission line fault and robot operation fault,and the probability of circuit inspection fault is calculated to realize the quantification of fault diagnosis results of robot circuit inspection.Compared with the traditional method,the average diagnostic accuracy of the proposed method was 93.4%,which increased by 10.8%.
作者
苗俊
袁齐坤
刘立文
尤志鹏
王璋
MIAO Jun;YUAN Qi kun;LIU Li wen;YOU Zhi peng;WANG Zhang(Qujing Power Supply Bureau of Yunnan Power Grid Co.,Ltd.,Qujing 655000,China)
出处
《电子设计工程》
2020年第9期184-188,共5页
Electronic Design Engineering
基金
2019年云南电网公司科技项目(050200K K52180004)。
