ZHANG Jiangfeng, YIN Feng, SU Ye, et al. LVDT faults auto-monitoring technology for turbine governing valve based on video processing[J]. Thermal power generation, 2017, 46(12): 18-23.
ZHANG Jiangfeng, YIN Feng, SU Ye, et al. LVDT faults auto-monitoring technology for turbine governing valve based on video processing[J]. Thermal power generation, 2017, 46(12): 18-23.DOI:
基于视频处理的汽轮机调节阀LVDT故障自动监控技术
摘要
线性可变差动传感器(LVDT)是汽轮机伺服系统中反馈调节阀开度的重要测量部件
对汽轮机的安全稳定运行具有重要作用。本文在计算机视觉的基础上
采用状态拟合、目标跟踪及模式识别等算法
提出了一种能够适应于工业现场的汽轮机调节阀LVDT故障自动监控技术
通过LVDT视频图像目标区域的角点检测
判断LVDT视频图像拉杆位移量
从而确定调节阀开度
结合现场热工仪表信息
判断调节阀LVDT是否出现故障。该技术能够弥补人工巡检的不足
主动、连续对汽轮机调节阀LVDT的运行工况进行监控
有效避免因调节阀LVDT断裂、脱落等故障发现不及时导致汽轮机运行工况失衡的发生。
Abstract
The linear variable differential transducer(LVDT) is an important measuring component used to feedback the opening of valves in steam turbine servo system
which plays an important role in safe and stable operation of the steam turbine.On the basis of video processing
by using state fitting
target tracking and pattern recognition algorithm
this paper proposes an auto-monitoring technology for LVDT faults occurred on turbine governing valves
which can be used in industrial plants.Through the corner detection for LVDT video image target area
this technology judges the LVDT video image rod displacement
so as to determine the opening of the control valve.Then
combining with the field thermal instrument information
it determines whether the control valve LVDT fails.The technology can replace manual inspection and monitor the valve LVDT initiatively and continuously
which can effectively avoid the imbalance of turbine operation caused by LVDT faults not be found timely