Problem
Research questions and friction points this paper is trying to address.
High-dimensional optimization
Expensive black-box problems
Efficient global optimization
Engineering design
KT-EGO: A Knowledge Transfer Assisted Efficient Global Optimization Algorithm for Solving High-Dimensional Expensive Black-Box Problems
🤖 AI Summary
This study addresses the challenges of the curse of dimensionality and prohibitive evaluation costs in high-dimensional expensive black-box optimization by proposing an efficient global optimization framework based on subspace decomposition and knowledge transfer. The method partitions the high-dimensional search space into lower-dimensional subsets, accelerating global exploration through cross-subspace knowledge transfer. Furthermore, it innovatively incorporates a surrogate-based data fusion strategy and an adaptive variable-range search mechanism to effectively balance exploitation and exploration. Experimental evaluations on standard benchmark functions and a compressor blade aerodynamic design task demonstrate that the proposed algorithm significantly improves both convergence efficiency and solution quality for high-dimensional expensive optimization problems.
📝 Abstract
Many engineering problems involve optimizing a high-dimensional expensive black-box (HEB) design space. To solve such problems efficiently, we propose a knowledge transfer assisted efficient global optimization (EGO) algorithm, labeled as KT-EGO, which extends the EGO algorithm for solving problems over higher dimensions (i.e., $d>20$). Specifically, the original design space is divided into several low-dimensional subset design spaces. More importantly, in order to extract information from the subset design spaces to accelerate the progress of full optimization, we propose a surrogate-based data fusion strategy in KT-EGO. And further, a searching strategy with an adaptive variable range is devised to enhance the exploitation of promising areas. To show the effectiveness of our proposed algorithm, it is compared against the state-of-the-art algorithms over 12 benchmark functions and a 28-dimensional engineering optimization for the design of compressor blade, which fully validates the effectiveness of the KT-EGO for solving HEB problems.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Knowledge Transfer
Efficient Global Optimization
High-Dimensional Expensive Black-Box Problems
Surrogate-based Data Fusion
Adaptive Variable Range
🔎 Similar Papers
No similar papers found.
