Sensitivity analysis of predictive models with an automated adjoint generator [electronic resource]
The adjoint method is a well established sensitivity analysis methodology that is particularly efficient in large-scale modeling problems. The coefficients of sensitivity of a given response with respect to every parameter involved in the modeling code can be calculated from the solution of a single...
Saved in:
| Online Access: |
Online Access |
|---|---|
| Corporate Author: | |
| Format: | Government Document Electronic eBook |
| Language: | English |
| Published: |
Oak Ridge, Tenn. : Oak Ridge, Tenn. :
Oak Ridge National Laboratory. ; distributed by the Office of Scientific and Technical Information, U.S. Department of Energy,
1987.
|
| Subjects: |
| Summary: | The adjoint method is a well established sensitivity analysis methodology that is particularly efficient in large-scale modeling problems. The coefficients of sensitivity of a given response with respect to every parameter involved in the modeling code can be calculated from the solution of a single adjoint run of the code. Sensitivity coefficients provide a quantitative measure of the importance of the model data in calculating the final results. The major drawback of the adjoint method is the requirement for calculations of very large numbers of partial derivatives to set up the adjoint equations of the model. ADGEN is a software system that has been designed to eliminate this drawback and automatically implement the adjoint formulation in computer codes. The ADGEN system will be described and its use for improving performance assessments and predictive simulations will be discussed. 8 refs., 1 fig. |
|---|---|
| Item Description: | Published through SciTech Connect. 01/01/1987. "conf-870859-19" "DE88000576" 9. annual low-level radioactive waste management program conference, Denver, CO, USA, 25 Aug 1987. Oblow, E.M.; Pin, F.G. |
| Physical Description: | Pages: 6 : digital, PDF file. |