Adaptive Echo State Network to maximize overhead power line dynamic thermal rating

A widely distributed power line sensor network (PLSN) has been proposed to monitor such a utility asset's status. One of its important applications is to evaluate the real time dynamic current capacity of overhead power lines down to `per span' level of granularity, and thus to maximize the existing power grid utilization. How to predict the conductor temperature ahead of time subject to various conductor overload conditions is the most critical and challenging task to evaluate the line dynamic thermal rating. This paper proposes an Echo State Network (ESN) to adaptively identify the nonlinear overhead conductor thermal dynamics under different weather conditions, and to predict the conductor temperature. This method requires only temperatures and line current as inputs and its simplified calculation makes it an attractive and cost effective solution to real-time implementation. Furthermore, by continuously providing accurate real-time line thermal conditions, this method can assist in utilizing the power lines more effectively.