基于贝叶斯决策树的电网报警信息去噪方法研究

电网报警信息中存在着大量的噪声数据,会严重影响事故处理的快速性和准确性。为了对电网报警信息进行噪声滤除,将报警信息与OMS中的检修记录以及SCADA中的遥测值相关联,滤除一部分错误报警信息,可以减少样本数量。针对决策树算法存在分类二义性的缺陷,采用贝叶斯决策树算法进行去噪。算例编程证明,该方法可以提高报警信息的去噪准确率。

Denoising Method of Grid Alarm Information Based on Bayesian Decision Tree

Grid alarm information exists a lot of noise data that will seriouslv affect the speed and the accuracy of accident treatment.In order lo filter the grid alarm noise information,the alarm information,maintenance record in OMS and Telemetry values in SCADA are associated to filter a part of error alarm information,which can reduce the number of samples.Aiming at the delect of decision tree algorithm,the Bayesian decision tree algorithm is applied for denoising.The example proves that this method can improve the accuracy of alarm information denoising.