..
Copyright 2021 IRT Saint Exupéry, https://www.irt-saintexupery.com
This work is licensed under the Creative Commons Attribution-ShareAlike 4.0
International License. To view a copy of this license, visit
http://creativecommons.org/licenses/by-sa/4.0/ or send a letter to Creative
Commons, PO Box 1866, Mountain View, CA 94042, USA.
..
Contributors:
:author: Francois Gallard, Damien Guenot, Charlie Vanaret
Glossary
--------
.. glossary::
:sorted:
MDA
Multi-Disciplinary Analysis
MDO
Multi-Disciplinary Design Optimization
MDF
Multi-Disciplinary Feasible
IDF
Individual Discipline Feasible
BLISS
Bi-Level Integrated System Synthesis,
BLISS98
Bi-Level Integrated System Synthesis, original BLISS variant from 1998 paper using coupled derivatives
BLISS2000
Bi-Level Integrated System Synthesis, BLISS variant from 2000 paper, using surrogate models
CO
Collaborative Optimization
ANN
Artificial Neural Network
CSSO
Concurrent Subspaces Optimization
MDOIS
MDO of Independent Subspaces
AAO
All At Once
ASO
Asymmetric Subspace Optimization
SAND
Simultaneous Analysis and Design
ATC
Analytical Target Cascading
WORMS
Workflow Management System
eWORMS
eclipse Workflow Management System
CFD
Computational Fluid Dynamics
CSM
Computational Structure Mechanics
EA
Evolutionary Algorithm
SBO
Surrogate Based Optimization
XDSM
eXtended Design Structure Matrix
LP
Linear Programming
LHS
Latin Hypercube Sampling
SOM
Self Organizing Map
API
Application Program Interface
COTS
Commercial Off The Shelf
SRS
Software Requirements Specification
HPC
High Performance Computing
UML
Unified Modeling Language
IRT
Institut de Recherche Technologique Saint Exupéry
DFO
Derivative-Free Optimization
ANR
Agence Nationale de la Recherche
KKT
Karun Kuhn Tucker
LOO
Leave One Out
RBF
Radial Basis Function
|g|
Generic Engine for Multi-disciplinary Scenarios, Exploration and Optimization
JSON
JavaScript Object Notation
XML
Extensible Markup Language
SSBJ
SuperSonic Business Jet
MMA
Method of Moving Asymptotes
SLSQP
Sequential Least Squares Quadratic Programming
FPO
Future Projects Office
HTML
Hypertext Markup Language
SWIG
Simplified Wrapper and Interface Generator
PIDO
Process Integration and Design Optimization
SLM
Simulation Life Cycle Management
DOE
Design Of Experiments
HDF
Hierarchical Data Format to save and structure files with huge data
grammar
A set of rules to be respected by a data set. Typically used to describe the inputs and outputs of a discipline
JSON schema
A JSON description of JSON data, similar to XML schemas
process
A series of executions and data exchanges, ie the workflow and the data flow.
workflow engine
A program used to design, run and analyze processes
processes data
The disciplines inputs and generated outputs during the execution of a process
serialization
Process of writing objects or data structures to disk or more generally, to formats that can be stored
work flow
The execution sequence of the disciplines in a process
data flow
The sequence of data creation by, and exchanges between disciplines in a process
MDO integrator
A user class that uses disciplines and MDO formulations to create, test and maintain MDO scenarios
disciplinary expert
A user class of the MDO platform that wraps disciplinary capabilities into disciplines and creates, tests or maintains disciplinary processes
MDO user
A user class of the MDO platform that executes an MDO scenario to produce results
MDO formulations designer
A user class of |g| that creates, implements, tests or maintains MDO formulations
algorithm integrator
A user class of the formulation engine that integrates mathematical algorithms such as an optimization algorithm, :term:`DOE` method or surrogate model
discipline
One program, or an arbitrary set of simulation software, that can be viewed as a mathematical multivalued function, taking inputs and producing outputs through its execution
chain
A process that executes a set of disciplines in a sequential way, where outputs of the previously executed disciplines are passed as inputs of the next ones
simulation software
A program that simulates a part of the physics of a system, or contributes to the overall simulation of the system, such as a mesher
MDO formulations engine
A program that enables the implementation of MDO formulations
wrapper
Here discipline wrappers. Standardized interface defining inputs, outputs and execution of a given simulation software.
library wrapper
A code that translates the existing API of a program, or a library, into a compatible one
interface
A set of functions and data from a software exposed to other software
design problem
An engineering problem such that a shape has to be changed to match or improve criteria under constraints
MDO formulation
The mathematical strategy used to define the optimization problem(s) to be solved
bi-level
A type of :term:`MDO formulation` which formulates multiple optimization problems
Monolithic
A type of :term:`MDO formulation` which formulates a single optimization problem
disciplinary optimization
The most basic MDO formulation restricted to a single set of design variables and only suited for a :term:`weakly coupled problem`
weakly coupled problem
A multidisciplinary problem where the coupling variables can be computed by a single execution :term:`chain`
MDO architecture
the software architecture that enables the programming and resolution of MDO design problems
scenario
The translation of a design problem into an executable. When executed, a scenario generates a :term:`process`
Design Of Experiments
A sampling of a design space, or a generic method that produces samplings of design spaces
design space
The mathematical set containing the design variables of an optimization problem
design variables
The unknowns of the optimization problem
system design variables
The design variables that are shared by more than one discipline, at the system level optimization problem in a bi level MDO scenario
shared design variables
The design variables that are shared by more than one discipline
coupling variables
In an MDO scenario, variables that are both used as inputs of a discipline and outputs of another one, or the same discipline
local design variables
In an MDO scenario, a subset of the design variables that are inputs of only one discipline
disciplinary design variables
In an MDO scenario, a subset of the design variables that are inputs of only one discipline
operating condition
An input parameter of a simulation program that defines a physical parameter in which the system operates. Typically the speed of a vehicle, or its altitude
MDO platform
A set of programs integrated in a common framework, enabling the resolution of multidisciplinary engineering problems using numerical simulation and optimization
generic process
A process that can be applied to any discipline or set of disciplines, such as :ref:`mda` methods that solves the coupling variables of a set of disciplines
optimization problem
A mathematical problem consisting in finding a set of variables which minimizes or maximizes a mathematical function (possibly a set-valued map), while satisfying constraints on these variables or on artrary functions
optimization history
The database of values of the objective function, constraints and design variables obtained during an optimization
optimization algorithm
An algorithm capable of solving optimization problems
DOE algorithm
An algorithm that generates samples of the design space
trade-off
A study that aims at comparing different options in terms of design parameters, and analyse their impacts
driver
A :term:`optimization algorithm` or :term:`DOE`
objective function
The function to be minimized or maximized in an optimization problem
constraint
A function of the design variables that must be kept either null or negative in an optimization problem
constraints
All the functions of the design variables that must be kept either null or negative in an optimization problem
surrogate model
A mathematical model of another model. Typically used to substitute an expensive simulation-based model by an approximation whose cost of evaluation is lower, at the price of an initial sampling of the original model
workflow-driven
Characteristic of a workflow engine, for which processes are described through the work flow, and in which the data flow is deduced accordingly
data-driven
Characteristic of a workflow engine, for which processes are described through the data flow, and in which the work flow is deduced accordingly
modular architecture
A software architecture based on separated components, with a relative independence between them
Model Center
COTS workflow engine developed by Phoenix Integration
Scilab
An open-source alternative to Matlab
LSF
COTS jobs scheduler for HPC clusters
NLopt
Non-Linear Optimization package, an open source library of optimization algorithms from MIT, http://ab-initio.mit.edu/nlopt
SciPy
Open Source Library of Scientific Tools, containing a library of optimization algorithms, https://www.scipy.org
Secure Shell
(SSH) an encrypted network protocol for accessing remote computers
sequence diagram
A :term:`UML` diagram that shows objects execution sequence, such as the function calls and their arguments, and the execution order
job scheduler
A program used to distribute tasks and allocate resources for tasks on HPC clusters
complex step
A numerical method to approximate the derivative of a function, similarly to finite differences, but using complex perturbations
finite differences
A numerical method to approximate the derivative of a function by small perturbations of the inputs
design document
A document that describes the design of a program, responding to a Software Requirements Specification
OpenDACE
Open Design and Analysis of Computer Experiments: a program developed by Airbus to standardize interfaces of optimization algorithms, :term:`DOE` methods and surrogate models as well as the related problems to be solved
gradient-based optimization
A class of optimization algorithms that use the total derivatives of the objective function and constraints
gradient
Total derivative of a function with respect to its variables
jacobian
The matrix of first order partial derivatives of outputs with respect to inputs
Hessian
The matrix of second order partial derivatives of one output with respect to inputs
optimum
Solution of an :term:`optimization problem`: the :term:`design variables` values at the minimum of the function, subject to the constraints
Object Oriented Programming
A programming paradigm based on objects, which are data structures as well as a structure for methods
run time
In computer science, run time, runtime or execution time is the time during which a program is running (executing), in contrast to other program life cycle phases such as compile time, link time and load time
pip
The PyPA recommended tool for installing Python packages. https://pypi.python.org/pypi/pip
anaconda
A python tool to create virtual environment and easily install precompiled packages, https://www.anaconda.com/distribution
fixed point
A family of numerical resolution methods based on an iterative sequence of execution of the type : A() -> B() -> C() -> ... -> A() -> B() -> C() until convergence
root finding
A family of numerical resolution methods that solves multivariate problems of the type R(x)=0
Newton method
A :term:`root finding` method that uses successive linear approximations of the function of interest