作者: Sharda Tripathi , Swades De
关键词:
摘要: This paper presents a novel data driven framework based on $\epsilon$ -Support Vector Regression to reduce the bandwidth requirement for transmission of phasor measurement unit (PMU) data. is achieved by judicious elimination redundant at PMU before transmission. Simultaneously, missing samples are predicted PDC ensure faithful identification impending disturbances in power system. Due inherent nonstationary nature data, hyperparameters dynamically recomputed as necessary, thereby maintaining accuracy prediction and robustness algorithm. Performance proposed algorithm evaluated via large scale simulations using powerline frequency A trade-off between quality runtime observed, which addressed suitable selection hyperparameters. Compared competitive reduction scheme, saves around 60% identifies system 73% more accurately.