Intelligent scheduling of source-grid-load-storage resources based on q-learning algorithm

In the fact of the increasing data transmission and computation demand in the source-grid-load-storage interaction scenario in the new electric power system, mobile edge computing has been proposed and widely used, which effectively reduces the delay and energy consumption of task computation by sinking the computation and storage resources to the edge of the network, and providing services for the users close to the mobile terminals. Therefore, for the offloading problem of the new power computing task in the cooperative interaction scenario of source-network-load-storage, a Q-learning-based decision-making method for offloading computing tasks is proposed, which comprehensively considers the network environment and the server state in the process of task offloading, takes the minimisation of the task latency and terminal energy consumption as the optimisation objective, and jointly optimises the latency, energy consumption and economic overheads generated by the offloading process. Simulation results show that the proposed scheme can effectively reduce the computational delay, energy consumption and total overhead of the system compared with other benchmark schemes.