Efficient parallel surrogate optimization algorithm and framework with application to parameter calibration of computationally expensive three-dimensional hydrodynamic lake PDE models

• Published in: Environmental modelling & software : with environment data news Vol. 135; p. 104910
• Main Authors: Xia, Wei, Shoemaker, Christine, Akhtar, Taimoor, Nguyen, Manh-Tuan
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.01.2021; Elsevier Science Ltd
• Subjects: Algorithms; Calibration; computer software; Derivatives; Dynamic coordinate search; Environment models; Environmental modeling; Evolutionary computation; Global optimization; Heuristic methods; High-performance computing; humans; hydrodynamics; Lake hydrodynamics; lakes; Optimization; Parallel optimization; Parameter calibration; Parameters; Surrogate model; system optimization; Three dimensional models; Global optimization; Surrogate model; Parameter calibration; Parallel optimization; High-performance computing; Dynamic coordinate search; Lake hydrodynamics
• ISSN: 1364-8152; 1873-6726
• DOI: 10.1016/j.envsoft.2020.104910

Parameter calibration for computationally expensive environmental models (e.g., hydrodynamic models) is challenging because of limits on computing budget and on human time for analysis and because the optimization problem can have multiple local minima and no available derivatives. We present a new general-purpose parallel surrogate global optimization method Parallel Optimization with Dynamic coordinate search using Surrogates (PODS) that reduces the number of model simulations as well as the human time needed for proper calibration of these multimodal problems without derivatives. PODS outperforms state-of-art parallel surrogate algorithms and a heuristic method, Parallel Differential Evolution (P-DE), on all eight well-known test problems. We further apply PODS to the parameter calibration of two expensive (5 h per simulation), three-dimensional hydrodynamic models with the assistant of High-Performance Computing (HPC). Results indicate that PODS outperforms the popularly used P-DE algorithm in speed (about twice faster) and accuracy with 24 parallel processors. •A new general-purpose parallel surrogate global optimization algorithm PODS.•PODS outperforms previous parallel methods on eight well-known test problems.•PODS defeats Differential Evolution on two expensive lake model calibration problems.•An open-source optimization toolbox including coupling PODS with Delft3D-FLOW.