Dakota V6.7 User's Manual

User Manual: Pdf

Open the PDF directly: View PDF .
Page Count: 364 [warning: Documents this large are best viewed by clicking the View PDF Link!]

Scroll down to view the document on your mobile browser.

Preface
Introduction
- Motivation for Dakota Development
- Dakota Capabilities
- Coupling Dakota to a Simulation
- User's Manual Organization
- Files Referenced in this Manual
- Summary
Dakota Tutorial
- Quickstart
- Dakota Input File Format
- Examples
- Dakota Command-Line Options
- Next Steps
  - Problem Exploration and Method Selection
  - Key Getting Started References
Parameter Study Capabilities
- Overview
  - Initial Values
  - Bounds
- Vector Parameter Study
- List Parameter Study
- Centered Parameter Study
- Multidimensional Parameter Study
- Parameter Study Usage Guidelines
- Example: Vector Parameter Study with Rosenbrock
Design of Experiments Capabilities
- Overview
- Design of Computer Experiments
- DDACE
- FSUDace
- PSUADE MOAT
- Sensitivity Analysis
  - Sensitivity Analysis Overview
  - Assessing Sensitivity with DACE
- DOE Usage Guidelines
Uncertainty Quantification Capabilities
- Overview
  - Summary of Dakota UQ Methods
  - Variables and Responses for UQ
- Sampling Methods
- Reliability Methods
- Stochastic Expansion Methods
  - Uncertainty Quantification Examples using Stochastic Expansions
    - Polynomial Chaos Expansion for Rosenbrock
    - Uncertainty Quantification Example using Stochastic Collocation
- Importance Sampling Methods
  - Importance Sampling Method based on Reliability Approach
  - Gaussian Process Adaptive Importance Sampling Method
- Adaptive Sampling Methods
  - Adaptive sampling based on surrogates
  - Adaptive sampling based on dart throwing
- Epistemic Nondeterministic Methods
  - Interval Methods for Epistemic Analysis
  - Dempster-Shafer Theory of Evidence
- Bayesian Calibration Methods
- Uncertainty Quantification Usage Guidelines
Optimization Capabilities
- Optimization Formulations
  - Constraint Considerations
- Optimizing with Dakota: Choosing a Method
- Additional Optimization Capabilities
  - Multiobjective Optimization
    - Multiobjective Example 1
    - Multiobjective Example 2
  - Optimization with User-specified or Automatic Scaling
    - Scaling Example
- Optimization Usage Guidelines
- Optimization Third Party Libraries
Nonlinear Least Squares Capabilities
- Overview
- Nonlinear Least Squares Fomulations
- Nonlinear Least Squares with Dakota
- Solution Techniques
- Examples
- Usage Guidelines
Models
- Overview
- Single Models
- Recast Models
- Surrogate Models
- Nested Models
- Random Field Models
- Active Subspace Models
Variables
- Overview
- Design Variables
  - Continuous Design Variables
  - Discrete Design Variables
- Uncertain Variables
  - Aleatory Uncertain Variables
    - Continuous Aleatory Uncertain Variables
    - Discrete Aleatory Uncertain Variables
  - Epistemic Uncertain Variables
    - Continuous Epistemic Uncertain Variables
    - Discrete Epistemic Uncertain Variables
- State Variables
- Management of Mixed Variables by Iterator
- Dakota Parameters File Data Format
  - Parameters file format (standard)
  - Parameters file format (APREPRO)
- The Active Set Vector
  - Active set vector control
Interfaces
- Overview
- Simulation Interfaces
- Building a Black-Box Interface to a Simulation Code
- Simulation Interface Components
- Simulation File Management
- Parameter to Response Mapping Examples
- Parameters and Results Files with dakota.interfacing
Responses
- Overview
- Dakota Results File Data Format
- Active Variables for Derivatives
Inputs to Dakota
- Overview of Inputs
  - Tabular Data Formats
- Data Imports
Output from Dakota
- Overview of Output Formats
- Standard Output
- Tabular Output Data
- Graphics Output
- Error Messages Output
- Stochastic expansion exports
- Surrogate Model Exports
- Variables Output from Pre-run
Advanced Methods
- Overview
- Hybrid Minimization
- Multistart Local Minimization
- Pareto Optimization
- Mixed Integer Nonlinear Programming (MINLP)
  - Example MINLP Problem
- Surrogate-Based Minimization
  - Surrogate-Based Local Minimization
  - Surrogate-Based Global Minimization
Advanced Model Recursions
- Mixed Aleatory-Epistemic UQ
- Optimization Under Uncertainty (OUU)
- Surrogate-Based Uncertainty Quantification
Advanced Simulation Code Interfaces
- Algebraic Mappings
- Developing a Direct Simulation Interface
- Existing Direct Interfaces to External Simulators
  - Matlab
    - Dakota/Matlab input file specification
    - Matlab .m file specification
  - Python
- Scilab Script and Direct Interfaces
Parallel Computing
- Overview
  - Categorization of parallelism
  - Parallel Dakota algorithms
- Single-level parallelism
- Multilevel parallelism
- Capability Summary
- Running a Parallel Dakota Job
  - Single-processor execution
  - Multiprocessor execution
- Specifying Parallelism
- Application Parallelism Use Cases
Restart Capabilities and Utilities
- Restart Management
- The Dakota Restart Utility
Simulation Failure Capturing
- Failure detection
- Failure communication
- Failure mitigation
- Special values
Additional Examples
- Textbook
  - Gradient-based Constrained Optimization
- Rosenbrock
  - Least-Squares Optimization
- Herbie, Smooth Herbie, and Shubert
  - Efficient Global Optimization
- Sobol and Ishigami Functions
- Cylinder Head
  - Constrained Gradient Based Optimization
- Container
  - Constrained Gradient Based Optimization
- Cantilever
  - Constrained Gradient Based Optimization
  - Optimization Under Uncertainty
- Multiobjective Test Problems
  - Multiobjective Test Problem 2
  - Multiobjective Test Problem 3
- Morris
  - Morris One-at-a-Time Sensitivity Study
- Test Problems for Reliability Analyses
- Test Problems for Forward Uncertainty Quantification

Navigation menu