多源数据融合下的电力系统数据传输数字化风险检测技术

Digital risk detection technology for power system data transmissionunder multi-source data fusion

多源数据通信过程中,主要依托于频域特征检测异常数据传输信号,忽略了时域特征,使得数据传输异常检测结果漏检率较高。因此,提出多源数据融合下的系统数据传输异常检测技术。考虑系统多源数据融合过程中的分布式数据传输特点,建立包含数据库、传感结合、数据存储安全构件的传输数据采集模型。获取每个通道的数据传输异常特征序列,并实现数据传输序列的重构处理。针对重构序列提取时域特征和频域特征,推算出不同数据传输周期的特征尺度关联系数,从而检测出数据传输异常通信流量。解析包含异常通信流量的数据流,建立DOM(文档对象模型)树结构,通过节点量化评估得出数据传输异常指数,输出风险检测结果。实验结果表明:新提出的数据传输异常检测技术应用后,所得结果的最大漏检率仅为8.05%,更准确地识别出多源数据传输风险问题。

In the process of multi-source data communication,it mainly relies on the frequency domain feature to detect the abnormal data transmission signal,and ignores the time domain feature,which makes the detection result of data transmission anomaly miss rate high.Therefore,the abnormal detection technology of system data transmission under multi-source data fusion is proposed.Considering the characteristics of distributed data transmission in the process of multi-source data fusion,a transmission data acquisition model including database,sensor combination and data storage security components is established.The data transmission anomaly sequence of each channel is obtained,and the data transmission sequence is reconstructed.The time-domain and frequence-domain features are extracted from the reconstructed sequences,and the feature-scale correlation coefficients of different data transmission cycles are deduced,so as to detect abnormal data transmission traffic.Analyze the data flow containing abnormal communication traffic,establish DOM(document object model)tree structure,obtain data transmission anomaly index through node quantitative evaluation,and output risk detection results.The experimental results show that after the application of the new data transmission anomaly detection technology,the maximum missed detection rate is only 8.05%,which can more accurately identify the risk of multi-source data transmission.