A Real-Time Synchrophasor Data Compression Method Using Singular Value Decomposition

The proliferation of phasor measurement units (PMUs) presents new challenges in archiving and processing large amounts of synchrophasor data which necessitates advanced data compression methods. This paper proposes a singular value decomposition (SVD)-based method for compression of synchrophasor data, including magnitude, phase-angle, and complex phasor. The proposed method includes a dimensionality evaluation and reduction technique and a real-time progressive partitioning algorithm. The proposed dimensionality reduction technique employs the measurement uncertainty of PMUs and introduces a threshold criterion on the signal-to-noise ratio (SNR) of SVD modes. Singular modes with high SNR are retained, and those dominated by measurement error are discarded to achieve a high compression ratio (CR) while preserving the critical information with adequate accuracy. The proposed progressive partitioning separates the data corresponding to normal and disturbance conditions by monitoring the dimensionality variations in real-time. The partitions containing the data of similar dimensionality are separately compressed to further improve the accuracy and CR. The performance of the proposed method is evaluated and benchmarked against state-of-the-art methods using both field and simulated PMU data. The results show that the proposed method provides high CR while accurately preserving the critical information of events and disturbances.

相量测量单元（PMU）的大量增加给大量同步相量数据的存档和处理带来了新的挑战，这就需要先进的数据压缩方法。本文提出了一种基于奇异值分解（SVD）的同步相量数据压缩方法，包括幅值、相角和复相量。所提方法包括一种维度评估和降维技术以及一种实时渐进分区算法。所提出的降维技术利用了PMU的测量不确定性，并针对SVD模式的信噪比（SNR）引入了一个阈值标准。保留高信噪比的奇异模式，丢弃受测量误差主导的模式，以便在以足够精度保留关键信息的同时实现高压缩比（CR）。所提出的渐进分区通过实时监测维度变化来分离对应正常和扰动情况的数据。对包含相似维度数据的分区分别进行压缩，以进一步提高精度和压缩比。使用现场和模拟的PMU数据对所提方法的性能进行了评估，并与现有最先进的方法进行了对比。结果表明，所提方法在准确保留事件和扰动关键信息的同时提供了高压缩比。