KADAL

Kriging for Analysis, Design optimization, And expLoration (KADAL) is Flow Diagnostics Lab is a powerful Python code developed by the Flow Diagnostics Lab. It offers a comprehensive suite of Bayesian Optimization tools, including a range of surrogate modeling methods, sampling techniques, and optimization methods. Our program incorporates various surrogate models, such as Ordinary Kriging, Regression Kriging, Polynomial Kriging, Composite Kernel Kriging, and Kriging with Partial Least Square. These models enable accurate predictions and efficient optimization in a wide range of applications. In the Bayesian optimization module, we provide the Single Objective Bayesian Optimization (SOBO) algorithm, which utilizes the Expected Improvement (EI) criterion. Additionally, we offer the Multi-Objective Bayesian Optimization (MOBO) module, which employs the Pareto efficient global optimization (ParEGO) and Expected Hypervolume Improvement (EHVI) algorithms. These modules enable efficient optimization in both single and multi-objective scenarios. To address uncertainty, our code includes an uncertainty quantification (UQ) module, allowing users to assess the reliability of their results. We also offer a global sensitivity analysis (GSA) module based on Sobol Indices, which helps identify the most influential input parameters. Furthermore, our reliability analysis module utilizes Active Kriging – Monte Carlo Simulation (AK-MCS) to assess the reliability of systems under uncertainty. It’s important to note that our code is continuously evolving, and we are actively working on incorporating more sophisticated methods to enhance its capabilities. If you are interested in this software, please contact my professor via our lab website Flow Diagnostics Lab.