Multi-surrogate multi-tasking optimization of expensive problems

作者：

Highlights：

• A global model and a local model are used to balance the exploration and exploitation.

• Two models are regarded as two related tasks and optimized simultaneously.

• Some best solutions in the archive are used in initialization to speed up the convergence.

摘要

•A global model and a local model are used to balance the exploration and exploitation.•Two models are regarded as two related tasks and optimized simultaneously.•Some best solutions in the archive are used in initialization to speed up the convergence.

论文关键词：Computationally expensive problems,Multi-tasking optimization,Surrogate models,Radial basis function

论文评审过程：Received 12 February 2020, Revised 5 May 2020, Accepted 12 July 2020, Available online 18 July 2020, Version of Record 28 July 2020.

论文官网地址：https://doi.org/10.1016/j.knosys.2020.106262