基于D-S证据理论与嵌入式传感器的监控与故障诊断

在水电站的故障监控与报警系统中一般都采用实时采集传感器数据并进行处理分析的方法,这不仅增加了通信的负担而且由于恶劣的电磁环境也容易造成数据的误传;为了克服这些缺点,提出了一种采用嵌入式传感器与数据融合理论相结合的故障监控与诊断的方案,并且在吉林丰满水电站的发电机线圈温度的监控系统中得到了实际应用

D-S Evidence Theory Applied to Fault Diagnosis of Hydropower Station Based on Embeded Sensors

Data fusion method is applied to the monitoring system of hydropower station. Because electromagnetic disturbance exists widely in the field, embedded multi sensors are used to aggregate initial data and bring out local judges in the data fusion architecture instead of simple multi sensors. In the fusion center, D S evidence theory is applied to the fusion algorithm to correspond with the embedded multi sensors. Moreover D S evidence theory need no precise understanding about the probability distribution of parameters of monitored system which is necessary while using Bayesian theory. By this kind of fusion architecture and algorithm, communication burden of monitoring system is decreased, thus result in lower error data transmitting rate and bring out more precise diagnosis. It has been applied to the monitoring system of Jilin Fengman Hydropower Station successfully.