A fast incremental Kriging-assisted algorithm for online optimization

doi:10.1016/j.asoc.2026.115695

A fast incremental Kriging-assisted algorithm for online optimization

2026-06-08

0

PRE

AI

DOI：10.1016/j.asoc.2026.115695

Original

Original request for help

Save

Abstract

En 中文

• A fast incremental Kriging-assisted algorithm is proposed to efficiently handle the online optimization problem of black-box functions. • A strategy for removing and adding new and old samples is proposed, and an incremental formula for the correlation matrix is derived to reduce the computational complexity of update to O((n-1)2). • The black-box function only needs to be calculated once for each time step in the time series data. • A total of four numerical cases, including time-varying systems, multiple local extrema and recursive systems, are used to verify the effectiveness of the proposed method. • The effectiveness of the proposed method is reduced due to the noise and model errors, but it still maintains good robustness.

A fast incremental Kriging-assisted algorithm for online optimization

Abstract

Journal

Organization