The role of learning techniques in synchrophasor-based Dynamic Thermal Rating

Abstract

Dynamic Thermal Rating is a key instrument to improve the exploitation of energy produced by renewable sources, since it allows to maximize the utilization of the existing transmission assets. The application of such techniques requires the reliable estimation of lines’ conductor temperature, which must be kept under specific threshold values in order to guarantee the safe operation of the whole system.The advent of Wide Area Monitoring Systems allowed the development of estimation methodologies which are able to infer useful pieces of information about lines’ thermal state through the exploitation of the correlation between conductor temperature and electrical parameters. In light of this, the paper proposes an innovative methodology based on Machine Learning techniques aimed at directly correlating the conductor’s temperature measurements to the synchrophasors acquired at the ends of the line. First experimental results on a real case application demonstrate the effectiveness of the proposed methodology in replicating measurements of temperature sensors installed in critical sections of transmission lines.