DRL-Based Online Task Offloading and Energy Resource Aggregation for Edge-Computing-Empowered Smart Grid Networks
Chuan Liu, Lei Chen, Wei Gao, Xi Zhang, Wei Peng, Feng Shu · IEEE Internet of Things Journal · 2024
The smart grid is expected to be integrated with advanced communication and control technologies to enhance its efficiency and reliability, enabling bidirectional information and energy exchanges between power providers and consumers. Considering the substantial data generated by the smart grid, mobile-edge computing (MEC) is introduced to address the need for considerable computing capacity. Limited attention has been directed toward exploring the application of MEC in smart grid scenarios. In this article, we innovatively consider maximizing the sum of computing rate and weighted energy trading benefit for a scenario with the distributed energy resource aggregation, smart grid, and MEC. We formulate the considered utility maximization as a mixed integer nonlinear programming (MINLP) problem. In order to solve this problem, we break it down into two distinct subproblems: 1) the binary offloading decision problem and 2) the energy allocation problem. We propose a deep reinforcement learning (DRL)-based framework to determine the offloading decision and design an optimization algorithm for the energy allocation problem. The simulation results indicate that the proposed algorithm achieves approximately 98% of the traversal algorithm’s performance with only one-thousandth of its processing time.