DRL-Assisted Topology Identification and Time Synchronization for PIoT-Empowered Distributed Renewable Resource Dispatch
Zhenyu Zhou, Chen Liu, Zhao Wang, Lei Lv, Le Zhang, Xing Li, Xiaoyan Wang, Wenwen Sun, Guoqing He, Yun Liu · IEEE Internet of Things Magazine · 2022
Power line communication-empowered power Internet of Things (PLC-PIoT) integrates the reliable PLC two-way automatic communication system (PLC-TWACS) and low-cost high-rate power line carrier communication (PLCC) to provide data collection, transmission, and processing for distributed renewable resource dispatch. Topology identification and time synchronization are two key technologies in PLC-PIoT, which still faces challenges such as unreliability of topology identification under frequent electrical equipment switching, large delay and low synchronization precision, as well as multi-timescale coupling under uncertain information. To address these challenges, we propose a Deep reinforcement learning (DRL)-assisted Topology Identification and tiMe Synchronization (DTIMS) framework to realize multi-timescale topology identification and time synchronization. Specifically, DTIMS combines PLCC and PLC-TWACS to achieve robust hierarchical topology identification in a large timescale, and explores DRL to achieve high-precision and low-delay intelligent time synchronization in a small timescale. Finally, DTIMS is validated through a case study to demonstrate its superior performance.