Optimization offers an integrated set of
powerful capabilities for single and
multi-attribute optimization. Through Design of
Experiments (DOE) and Response Surface Modeling
(RSM) techniques, engineers gain a rapid insight
in all the possible design options that meet
their requirements. Using advanced Optimization
routines, Virtual.Lab automatically selects the
optimal design, taking into account its
sensitivity to real-world variability, and
meeting the strictest robustness, reliability
and quality criteria.
Process Automation
This module allows users to define design
objectives, the design parameters and their
distribution. It automates the Optimization
loops by monitoring the whole process and by
eliminating tedious trial-and-error tasks in
engineering design analysis and process
planning. “Virtual” experiments are submitted
and the results can be visualized through
various Response Surface Modeling techniques.
Design Space Exploration
Design Space Exploration allows the user to
automatically explore and visualize the design
space through the use of a wide variety of
Design Of Experiments techniques. The results of
these experiments (such as parameter
contributions, design sensitivities and
correlation plots) give critical insight into
the variations of a design and the related
‘trade-offs’.
Optimization – NLP
The Optimization – NLP module allows users
to intelligently drive and optimize LMS
Virtual.Lab simulation processes. An optimal
design can be found using a variety of
optimization methods based on Non-Linear
Programming techniques, including Sequential
Quadratic Programming and Generalized
Reduced Gradient methods.
Optimization – Global
With the Global Optimization module,
three different state-of-the-art
algorithms - Differential Evolution
(DE), Self-adaptive Evolution (SE) and
Simulated Annealing (SA) - are available
for solving general constrained
optimization problems. The above
algorithms have a high probability of
efficiently finding a global optimum.
Optimization – Discrete
Discrete Optimization solves general
constrained optimization problems
including a mixture of continuous
and discrete variables. Some design
parameters can only take integer
values or can be chosen from a
limited list, such as for sheet
metal thicknesses. A choice of
discrete variables is available:
either integer values only, or a
catalog of real values, or a list of
strings. Special optimization search
routines effectively take into
account the discrete character of
the input variables.
Robust Design and Variability
In today’s highly competitive world
an optimal design is pushed close to
the design limits. Due to tolerances
on the design parameters, a response
cannot be considered as a single
deterministic value. Input
parameters are to be considered as
distributions instead. As a
consequence, any variation of a
design around its optimal value has
to be evaluated in order to meet the
desired robustness, reliability and
quality criteria.
Based on OPTIMUS from NOESIS Solutions
LMS Virtual.Lab Optimization is
based on the functionality of
OPTIMUS from NOESIS Solutions, a
subsidiary company of LMS
International. OPTIMUS is an
open software environment for
capturing and automating
analysis processes and
design-space navigation. OPTIMUS
empowers the intuitive capturing
and automation of any data or
simulation-based design process.
Design drivers, such as
Design-Of-Experiments (DOE),
multidisciplinary design
optimization and
quality-engineering techniques,
give more insight in the
behavior of designs and provide
better and higher-quality
designs faster than before.
[Back]
