基于模式识别的WAMS有功功率错误数据处理

发布时间：2018-03-29 02:03

本文选题：广域测量系统　切入点：模式识别　出处：《电网技术》2017年03期

【摘要】：目前,国内广域量测系统(WAMS)在实际应用中出现较多错误数据,由此衍生出大量错误的电网设备告警记录。为解决传统方法在辨识和还原WAMS错误数据时时效性差、漏报率高的问题,文章重点分析了在有功功率量测过程中所出现的异常情况,归纳和总结出典型错误数据的特征,提出一种基于模式识别的WAMS错误数据快速辨识和恢复方法。该方法根据数据特点提前设定错误数据和正常数据的标准特征向量,并将实测数据以同样方法转换成判断向量,从而进行模式匹配,实现错误数据的辨识。最后文中设计了3个算例分别展示了暂态下正确数据、稳态下错误数据以及大量数据下的辨识和恢复结果。结果表明,因为无需输入大量网架信息和没有迭代计算,模式识别的方法无论是准确度还是计算时间上都具有较强优势,能够为电网状态监测和分析等提供较可靠的数据支持。
[Abstract]:At present, there are more error data in domestic wide area measurement system (WAMS), which gives rise to a large number of wrong warning records of power network equipment. In order to solve the time-efficiency difference of traditional methods in identifying and restoring WAMS error data, The paper analyzes the abnormal situation in the process of active power measurement and summarizes the characteristics of typical error data. In this paper, a method for fast identification and recovery of WAMS error data based on pattern recognition is proposed, which sets the standard eigenvector of the error data and normal data in advance according to the characteristics of the data, and converts the measured data into a judgment vector by the same method. Finally, three examples are designed to show the correct data in transient state, the error data in steady state and the results of identification and recovery under a large number of data. Because there is no need to input a large amount of grid information and there is no iterative calculation, the method of pattern recognition has a strong advantage in both accuracy and computing time, and can provide reliable data support for power grid state monitoring and analysis.
【作者单位】：华南理工大学电力学院;
【分类号】：TM76

【参考文献】