The “IOSO Approximation” software allows the user to create fit
functions using powerful univariate and multivariate fitting
methods.
One-dimensional approximation methods:
- Polynomial Interpolation (PI)
- Polynomial Fitting (PF)
- Rational Interpolation (RI)
- Rational Fitting (RF)
- Linear Spline (SLI)
- Cubic Spline (SCI)
- Catmull-Rom Spline (SCRI)
- Akima Spline (SAI)
- Cubic Spline Fitting (SCF)
- Hermite Spline Fitting (SHF)
- Penalized Spline Fitting (SPF)
Multi-dimensional approximation methods:
- RBF - Radial basis function interpolation;
- IDW Inverse distance weighting interpolation.
The initial Data for the construction of the approximation
function can be:
-
protocols of projects of the IOSO optimization software;
- arrays of data in text format files, including CSV files;
- data entered into the IOSO Approximation table.
The
constructed approximation function can be saved as an executable
model (exe file) and used later as surrogate models in problems
of modeling systems or objects. These models can be used in the
IOSO optimization software to replace complex resource-intensive
computational models when solving optimization problems with
multidisciplinary projects and the possibility of automated
creation, connection and training of approximation models on new
data is implemented.
