模糊证据推理下的智能变电站运行态势异常实时在线告警仿真

智能变电站典型场景动态联动过程中，运行态势细节特征单一且冗余量大，为及时发现运行态势异常，提出模糊证据推理下智能变电站运行态势异常实时在线告警方法。利用模糊证据推理每种特征向量下隶属度值，归一化处理，获得整个所有隶属度值，依据判决准则对告警样本实施时间化处理，缩小告警范围，挖掘运行态势异常特征，预测出智能变电站可能发生的异常运行态势和故障，凭借Tdengine电力运维平台自动发出告警。实验结果表明，所提方法的运行态势异常识别率高、态势分析耗时短、加入噪声后的告警时延短。明确各个设备间的关联性，当其中一个环节出现问题，及时发出告警信息，避免产生连锁问题。

Real-time Online Alarm Simulation of Abnormal Operation Status of Intelligent Substation Under Fuzzy Evidence Reasoning

In the process of dynamic linkage of typical scenes of intelligent substation, the details of operation situation are single and redundant. In order to find the abnormal operation situation in time, a real-time online alarm method for abnormal operation situation of intelligent substation under fuzzy evidential reasoning is proposed. Fuzzy evidence is used to infer the membership value under each feature vector, normalize it, and obtain all the membership values. According to the judgment criteria, the alarm samples are processed in time to narrow the alarm range, mine the abnormal operation situation, predict the abnormal operation situation and faults that may occur in the intelligent substation, and automatically send an alarm by virtue of the Tdengine power operation and maintenance platform. The experimental results show that the proposed method has high recognition rate of abnormal situation, short time-consuming situation analysis, and short warning delay after adding noise. Clarify the relevance between various equipment. When one of the links has problems, send out alarm information in time to avoid interlocking problems.