Optimization
optiSLang offers efficient methods for solving various multidisciplinary, nonlinear and multicriteria optimization problems.
Method overview
Approximation of the objective function
- Gradient-based optimization (NLPQLP, L-BFGS)
- Evolutionary algorithms (genetic algorithms, evolution strategies)
- Pareto optimization
- Global response surface methods (linear, quadratic, moving least square)
- Adaptive response surface method (ARSM)
Features
Optimization of a side barrier shape regarding dummy injury criteria
- Continuous, discrete and binary design variables (up to many thousands)
- Definition of desing boundary conditions or manufacturing constraints
- Constraints of equality and inequality (parameter- free constraint handling)
- Robust defaults of algorithmic parameters
Focus
Calibration of measurement and simulation as optimization problem
The calibration of measurement and computation is one of the classic problems of model validation. If the difference between measured and computed data is too large, an optimization problem for minimizing the difference can be formulated. Optimization problems of calibration of measurement and computation are often also called identification problems.
optiSLang can be used for an automated identification procedure by performing a sensitivity analysis followed by an evolutionary algorithm. Practical experience has shown that a successful optimization problem can often only be formulated, if the design space can be assembled from sensitive parameters.
Postprocessing
Postprocessing window
The results of the optimization are visualized by an interactive postprocessing. For all parameters and responses, histories can be generated and individual designs can be selected.
Distributors
optiSLang is sold worldwide. A list of our distributors can be found here.
Your Contact
