A dynamic granularity-based multi-objective evolutionary algorithm for coal mine integrated energy system dispatch optimization

•A dynamic granularity search (DGS) approach is introduced to reduce the dimensionality of large-scale optimization problems.•A diversity-enhanced environmental selection (DES) strategy is designed to maintain both global and local diversity.•Integrating DGS and DES, a new large-scale constrained mu...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Expert systems with applications Ročník 296; s. 129142
Hlavní autoři: Zhong, Xiaoyu, Yao, Xiangjuan, Qiao, Kangjia, Gong, Dunwei
Médium: Journal Article
Jazyk:angličtina
Vydáno: Elsevier Ltd 15.01.2026
Témata:
Coal mine integrated energy system
Dynamic dispatch granularity
Large-scale
Multi-objective dispatch optimization
Strong constraints
Coal mine integrated energy system
Strong constraints
Large-scale
Multi-objective dispatch optimization
Dynamic dispatch granularity
ISSN:0957-4174
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:•A dynamic granularity search (DGS) approach is introduced to reduce the dimensionality of large-scale optimization problems.•A diversity-enhanced environmental selection (DES) strategy is designed to maintain both global and local diversity.•Integrating DGS and DES, a new large-scale constrained multi-objective evolutionary algorithm named DGMA is developed. The dispatch optimization of coal mine integrated energy system (CMIES) requires energy configuration at each time period (the smallest dispatch unit), which means the dimensionality of the variable space grows exponentially as the total number of time periods increases. Existing constrained multi-objective evolutionary algorithms converge slowly and struggle to identify tiny and disconnected feasible regions when solving the large-scale CMIES dispatch optimization problem. To this end, this paper develops a dynamic dispatch granularity-based multi-objective evolutionary algorithm, termed DGMA, where the dispatch granularity refers to the number of time periods represented by one dispatch unit. In particular, at the beginning of the evolution, the dispatch granularity is coarse (one dispatch unit denotes a large number of periods), which greatly reduces the search space and allows the proposed DGMA to rapidly detect potentially good solutions. As the evolution progresses, the dispatch granularity gradually refines, until one dispatch unit represents a single period. Through this gradual refinement, DGMA can perform a more refined search, thereby obtaining higher-quality dispatch plans. Furthermore, a diversity-enhanced environmental selection strategy is designed, which integrates the partition idea into the constraint relaxation method to promote both global and local diversity. The effectiveness and superiority of the proposed DGMA are validated through comparisons with seven state-of-the-art algorithms on various CMIES dispatch optimization problems.
ISSN:0957-4174
DOI:10.1016/j.eswa.2025.129142