Co-design of an unmanned cable shovel for structural and control integrated optimization: A highly heterogeneous constrained multi-objective optimization algorithm

Unmanned cable shovel is an intelligence-based large machine used for excavating in open-pit coal mines, with significant implications for the security and development of energy resources. The co-design of unmanned cable shovels, considering elements from both structural and control systems, has dem...

Full description

Saved in:

Bibliographic Details
Published in:	Applied energy Vol. 376; p. 124325
Main Authors:	Pang, Yong, Hu, Zhengguo, Zhang, Shuai, Guo, Guanchen, Song, Xueguan, Kan, Ziyun
Format:	Journal Article
Language:	English
Published:	Elsevier Ltd 15.12.2024
Subjects:	Co-design Constrained multi-objective optimization Expected hypervolume improvement Heterogeneous evaluations Kriging surrogate model Unmanned cable shovel Constrained multi-objective optimization Expected hypervolume improvement Kriging surrogate model Co-design Heterogeneous evaluations Unmanned cable shovel
ISSN:	0306-2619
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	Unmanned cable shovel is an intelligence-based large machine used for excavating in open-pit coal mines, with significant implications for the security and development of energy resources. The co-design of unmanned cable shovels, considering elements from both structural and control systems, has demonstrated superior performance in obtaining optimal solutions compared to the traditional approaches that focus on individual systems. However, the inherent complexities of diverse systems often lead to heterogeneous function evaluations, posing challenges for the existing optimization algorithms. To address highly heterogeneous multi-objective optimization problems characterized by surrogate-approximated expensive functions and explicit inexpensive functions, this study proposes a highly heterogeneous constrained multi-objective optimization (HHCMO) algorithm. Leveraging NSGA-III as an evolutionary optimizer and the Kriging model for approximating expensive functions, HHCMO strategically addresses the challenges posed by highly heterogeneous objective evaluations. A Monte Carlo sampling-based expected hypervolume improvement (EHVI) criterion, employing a new nearest point approximation method in a unified sampling space, effectively handles the impact of heterogeneous objective evaluations on the sequential infill of the Kriging model. Thorough evaluations on both unconstrained and constrained multi-objective benchmark problems validate the correctness of HHCMO and its superiority relative to state-of-the-art algorithms. Finally, the effectiveness of the proposed method is successfully demonstrated in an engineering scenario of an unmanned cable shovel, where it adeptly handles expensive functions from the structural system and inexpensive functions from the control system. [Display omitted] •The co-design of an unmanned cable shovel is studied.•The heterogeneity of the structural and control systems is considered.•A highly heterogeneous constrained multi-objective optimization method is proposed.•A sampling-based expected hypervolume improvement infill criterion is proposed.•The method is more appropriate to the highly heterogeneous optimization problem.
ISSN:	0306-2619
DOI:	10.1016/j.apenergy.2024.124325