This paper envisions a new control architecture for the protective relay setting in future power distribution systems. With deepening penetration of distributed energy resources at the end users level, it has been recognized as a key engineering challenge to redesign the protective relays in future distribution systems. The key technical difficulty lies in how to set up the control logic of relays so that they could accurately detect faulty conditions. The performance of traditional protection settings are limited by insufficient fault current either due to current limit of power electronics or high fault impedance. This paper proposes a new nested deep reinforcement learning approach to take advantage of the structural property of distribution networks and develops a new set of training methods for tuning the protective relays.
展开▼