Closed-loop integrated prediction-and-dispatching framework for unit commitment in power system with renewables and hydrogen energy

Yi Zhang, Xiaoming Zhang and Kejin Jia

Energy, 2025, vol. 329, issue C

Abstract: Wind and photovoltaic power coupling hydrogen is an effective way to improve the accommodation of renewable energy. As an important task of power system operation, unit commitment needs to consider not only randomness of renewables, but also the operation characteristics of electrolyzer and fuel cell. The traditional model for unit commitment problem is based on predict-then-dispatch implemented in a queue. The relevant dispatching information (i.e. cost, objective, constraint) is not fed back to the accuracy-oriented prediction model. Models with similar prediction accuracy can achieve different economic performance. Collaborative framework for integrating prediction and dispatching is proposed in this paper, which forms a closed loop structure to reduce the system cost. Economics evaluation model contains the income from hydrogen sale and operation cost of hydrogen subsystem. In addition to conventional constraints, dispatching optimization model includes not only network constraint of power flow security, but also the operational exclusivity of electrolyzer and fuel cell. MILP solver for optimal dispatching cannot efficiently deal with the computational burden, which is caused by the rapid increase of decision variables and constraints due to many historical scenarios. Therefore, data-driven machine learning method is introduced into the optimization problem. The cost-oriented prediction model is trained by the presented adaptive neuro-fuzzy interference system. How to select statistical and temporal characteristics of time series as model inputs has an impact on the performance of proposed integration framework. Furthermore, this paper discusses in detail the relationship between the model structure and system cost. Case study on the modified IEEE 24-bus power system demonstrates that the integrated prediction-and-dispatching framework can availably improve economics compared with traditional model.

Keywords: Integrated prediction-and-dispatching; Adaptive neuro-fuzzy interference; Cost-oriented; Feature selection; Economics evaluation; Wind/photovoltaic power coupling hydrogen (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0360544225021176
Full text for ScienceDirect subscribers only

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:eee:energy:v:329:y:2025:i:c:s0360544225021176

DOI: 10.1016/j.energy.2025.136475

Access Statistics for this article

Energy is currently edited by Henrik Lund and Mark J. Kaiser

More articles in Energy from Elsevier
Bibliographic data for series maintained by Catherine Liu ().