Multi-Energy Scheduling of an Industrial Integrated Energy System by Reinforcement Learning-Based Differential Evolution

Abstract

The Industrial Internet of Things (IIoT) is one of the main catalysts towards the realization of the Industry 4.0 paradigm, thus it is regarded as an essential element in future industrial systems - which can assist in reducing energy consumption and in enhancing product life-cycle management. In this study, an industrial multi-energy scheduling framework (IMSF) is proposed, with the aim of optimizing the usage of renewable energy and reducing the energy costs. The proposed method addresses the management of multi-energy flows in industrial integrated energy systems - incorporating multi-energy storage, renewable energy generation, energy conversion, and energy trading in a synchronous manner. The method considers the typical energy load of the industrial users, the energy price of the national grid and the trading platform, and the trade-off between investment costs and benefits from the various sub-systems. As this results in a complex system of systems, an artificial intelligence method is proposed to treat the problem, using reinforcement learning based differential evolution (RLDE), that can determine the optimal mutation strategy and associated parameters in an adaptive way. Case studies on real-world data evidence the effectiveness of the IMSF and the RLDE algorithm in reducing the energy costs in industrial environments.