Surrogate Based Design Optimization Toolbox
MATLAB Toolbox for metamodeling and solving optimization problem. The function to surrogate and optimize can be evaluated from an external numerical simulator.
SBDOT Copyright (C) 2017 CEA - LETI, DOPT. Author: C. Durantin
General informations
Metamodeling features
- Kriging (gaussian process based surrogate model)
- Radial Basis Function
- Co-kriging (multifidelity)
- CoRBF (multifidelity)
- BQQV Kriging (for qualitative variables)
Optimization features
- CMAES (constrained single objective)
- NSGA-II (constrained multiobjective)
- MGDA (unconstrained multiobjective)
Adaptive Sampling strategy
- Expected improvement (constrained global optimization)
- Gutmann criterion (constrained global optimization)
- Expected hypervolume improvement (multiobjective)
- Robust efficient global optimization
MATLAB toolboxes requirements
- Optimization toolbox
- Statistics toolbox