Hybrid deterministic procedure for automated sensitivity and uncertainty analysis
Conference
·
OSTI ID:5404867
A deterministic procedure for automating sensitivity and uncertainty analysis for large-scale studies is presented. The procedure makes use of the GRESS (GRadient Enhanced Software System) methodology to calculate the required sensitivities. GRESS is a precompiler that automatically enhances computer codes with computer calculus in such a way that partial derivatives can be calculated along with the normal results of the code. From these derivatives any sensitivity of interest can readily be obtained. The sensitivities can then be used in an uncertainty analysis procedure based on the quantification of covariances and parameter correlations, in conjunction with conventional statistical uncertainty methods. When this approach is used, the procedure may also include the use of a generalized least-square adjustment methodology for reducing the uncertainties in the final results by combining calculated or measured integral experiment data with the data used in the calculations. 7 refs.
- Research Organization:
- Oak Ridge National Lab., TN (USA)
- DOE Contract Number:
- AC05-84OR21400
- OSTI ID:
- 5404867
- Report Number(s):
- CONF-860905-8; ON: DE86013271
- Country of Publication:
- United States
- Language:
- English
Similar Records
An automated sensitivity analysis procedure for the performance assessment of nuclear waste isolation systems
Automated procedure for calculating time-dependent sensitivities in ORIGEN2
GRESS, a preprocessor for sensitivity studies on FORTRAN programs
Journal Article
·
Tue Dec 31 23:00:00 EST 1985
· Nucl. Chem. Waste Manage.; (United States)
·
OSTI ID:6984655
Automated procedure for calculating time-dependent sensitivities in ORIGEN2
Technical Report
·
Tue Oct 01 00:00:00 EDT 1985
·
OSTI ID:6225779
GRESS, a preprocessor for sensitivity studies on FORTRAN programs
Conference
·
Mon Dec 31 23:00:00 EST 1990
·
OSTI ID:6251397