基于WAMS量测数据的电网扰动和操作类型识别

摘要

The power disturbance and operation identification based on the measurements of WAMS(Wide Area Measurement System)and pattern recognition theory is introduced,including tie-line based region judgment,measurement pretreatment,pattern feature extraction and pattern matching. The feature extraction methods are summarized and based on the analysis of power disturbance and operation features more characteristic parameters are adopted in feature extraction to improve the identifiability. Simulative case study including various influencing factors,such as transition resistance,generator tripping ratio,load shedding ratio and so on,is carried out for a regional power grid. Results show its high accuracy,except for the short circuit with high transition resistance. Verification with the WAMS measurements of a regional power grid indicates the identification correctness rate is 100 ％,even for the secondary disturbances.%阐述了基于广域测量系统(WAMS)量测数据和模式识别方法进行电网扰动和操作类型识别的过程,主要包括基于联络线的扰动和操作区域判定、WAMS量测数据的预处理、模式特征的提取以及模式匹配.在模式特征的提取过程中,针对分析得到的不同类型扰动和操作的特征,采用了更多的电气特征量,并总结了相关的特征提取方法,提高了可识别性.采用某区域电网仿真算例对所述方法进行验证.并对影响扰动类型识别的多种因素如短路过渡电阻,切机、切负荷比例等进行了仿真研究,结果表明该方法均具有很高的正确率,只是对高阻短路的识别效果不够理想.应用实际扰动过程中电网WAMS量测数据进行验证,结果全部识别正确,且连锁扰动也能正确识别.