Multi-Objective energy management of Solar-Powered integrated energy system under forecast uncertainty based on a novel Dual-Layer correction framework

Abstract

Integrated energy systems (IESs) are increasingly pivotal in the global shift towards sustainable energy frameworks. Within IESs, the energy management system (EMS) plays a critical role, tasked with optimizing energy allocation to achieve objectives like grid stability, energy reliability, and cost-efficiency. A significant challenge for EMS is the inherent unpredictability of renewable energy. Given this, a model predictive control (MPC)-based real-time energy management framework is proposed, which aims to mitigate the impacts of radiation forecast uncertainties in solar-powered IES. A dual-layer correction mechanism is proposed to quantify forecast uncertainty, resulting in uncertain intervals inferred from the hidden Markov model (HMM). Then, the uncertain intervals are adeptly managed within the MPC decision-making framework through the constraint-tightening method. A case study on solar-powered electricity, heating, and hydrogen IES demonstrates the validity of the proposed framework. Five different sets of energy management strategies, such as deterministic analysis, distributionally robust optimization, and the proposed HMM-MPC are subjected to long-term simulations. Comparative analysis reveals that the proposed method enhances various aspects of the performance metrics, including demand response, energy reliability, and system self-sufficiency.