CENTAUR (Collection of ENgineering Tools for Analyzing Uncertainty and Reliability) is a software library that contains an array of methods for solving various types of problems, with an emphasis on non-deterministic analysis. CENTAUR provides methods for reliability analysis, distribution fitting, uncertainty quantification, numerical optimization, and more. CENTAUR has been under active development since 2008 and currently provides several of the reliability analysis methods used by NESSUS.
SwRI has used CENTAUR to develop custom software tools and solutions using the following platforms:
- Fortran
- C/C++
- MATLAB
- Python
- Microsoft Excel
CENTAUR Capabilities
- Reliability Analysis
- Sampling methods (Monte Carlo, LHS)
- MPP-based method (FORM, SORM, AMV, AMV+)
- Surrogate-based methods (EGRA)
- Hybrid methods (Multimodal adaptive importance sampling)
- Distribution Fitting
- 15 probability distribution models, including non-parametric modeling via kernel density estimation
- 3 fitting methods
- Uncertainty quantification via Bayesian inference, bootstrap resampling, and Fisher information
- Optimization
- Support for multiple, nonlinear constraints
- Gradient-based optimizers provided through OPT++
- Global optimization via DIRECT and EGO
- Surrogate Modeling
- Fit complex, nonlinear functions with state of the art Gaussian Process model
- Much more...