Application of a Real-Time Data Compression and Adapted Protocol Technique for WAMS

This paper proposes a real-time data compression and adapted protocol technique for wide-area measurement systems (WAMS). The compression algorithm combines exception compression (EC) with swing door trending (SDT) compression. The compression logic is designed to perform this algorithm in real time. Selection of compression parameters and data reconstruction are presented. An adapted protocol is introduced by improving the format of data frames defined by IEEE standard C37.118 for compressed data packets. The proposed compression technique and protocol were applied to the phasor measurement units (PMUs) of a hydropower plant in Guizhou Power Grid in Southwest China. A low-frequency oscillation incident was recorded by this technique. The raw, compressed and reconstructed data were analyzed to verify the compression and determine the accuracy of the proposed technique. Also, the wavelet-based data compression, the standalone EC and SDT are compared with the proposed compression technique. Our results demonstrated that this compression can reach the compression ratios in the range of 6 to 11. Also, this compression and adapted protocol technique can reduce the size of data packets by approximately 75% with high accuracy in both dynamic and steady states.

I. Introduction

Wide-Area measurement systems (WAMS) are becoming one of the most important monitoring systems for power systems. In contrast to supervisory control and data acquisition (SCADA) systems, WAMS provide synchronized phasor measurements with considerably higher reporting rates. The implementation of WAMS over the past 10 years has demonstrated their value for dynamic monitoring, system modeling and validation, closed-loop control, and wide-area protection [1]. However, their high reporting rates and numerous data channels result in large amounts of data, which need to be transported in communication systems and stored in control centers, measuring units, and application servers in real time [2], [3].