Real-Time Early Warning Study of Abnormal Monitoring Data in Flexible DC Converter Stations
To improve the early warning performance of abnormal data and maintain the operational stability of the flexible direct current(DC)converter station,a real-time early warning method for abnormal monitoring data in flexible DC converter stations is proposed.Firstly,the empirical rules are utilized to eliminate the data collision problem,and the abnormal data in it are accurately detected according to the abnormal values of the data information Hash function.Then,the wavelet transform is used to preprocess the abnormal monitoring data and extract key variable information from the abnormal data.Finally,according to the probability of abnormal data in the monitoring process,the early warning threshold is set,and then the early warning function is used to construct a real-time warning model.The experimental results show that after applying the method,the false alarm rate is controlled below 20%and the average absolute error is controlled below 0.12.Compared with the traditional method,the early warning results of this method are more effective and the accuracy is higher.
Flexible direct current(DC)converter stationMonitoring dataAbnormal dataData collisionScale parameterData detectionThreshold setting
NETL
NSTL
万方数据
