July 13, 2015
SBO algorithms for expensive functions optimization
There is a unique family of methods for optimization in pSeven Core embedded in pSeven - Surrogate-Based Optimization (SBO) or optimization based on approximation models. This approach allows to deal with computationally expensive problems, since optimization algorithm acts on surrogate model instead of direct functions evaluations.
Here is just a short list of SBO advantages:
- Is purely global optimization method
- Is insensitive to noisy responses
- Explicitly supports linear and quadratic constraints
- Allows to set a budget (number of function evaluations)
All this make SBO perfect for engineering optimization.
To build surrogate models for optimization purpose Optimizer block uses GP (Gaussian processes) technique. In addition to function prediction GP model can estimate the accuracy of the approximation model. This is crucial for SBO algorithms: search of the global optimum of the function turns into search of the best probability to discover global optimum. On pic 1. you can see the evolution of surrogate model while SBO is discovering the global optimum.
Pic. 1. Evolution of the surrogate model (green surface) while SBO is working. True function is blue surface
The outline of the algorithm is as follows (see pic. 2):
- Evaluate Design of Experiment;
- Build initial approximation model;
- Iteratively improve the approximation model evaluating the best candidates for global optimum.
Pic. 2. Typical Surrogate-Based Optimization history
Along with single-objective optimization SBO in pSeven Core supports multi-objective problems. In this case Multi-Objective Surrogate-Based Optimization algorithm works. MSBO can efficiently handle complicated optimization problems with tens of goal functions and hundreds of parameters.
By Dmitry Khominich, Head of Application tasks Department