Today is 2020-07-14 Tuesday,Welcome to this site 



Word:[Big][Middle][Small] QR Code 2018/6/6     Viewed:    

LS-OPT® for Design Optimization and Parameter Identification

        LS-OPT® is a simulation-based optimization tool which enables the solution of complex, multi-stage design processes or regression/classification tasks. LS-OPT interfaces with LS-DYNA® (e.g. result extraction) and also supports popular pre- and post-processors, e.g. for shape optimization. For visualization of results, graphical pre- and post-processing tools are included in the package.


Multidisciplinary and Multi-Objective Optimization (MDO/MOO)

Discrete and Mixed Optimization

Global Optimization

Robust and/or Reliability-based Optimization

LS-DYNA statistics, including outlier analysis and LS-PrePost® support

Parameter Identification with matching of noisy, steep and hysteretic curves

Full-field calibration using Digital Image Correlation

Uncertainty Quantification

Sensitivity Analysis

Solvers and Methods:

Sequential Response Surface Method

Genetic Algorithm and Efficient Global Optimization (EGO)

NSGA-II algorithm for MOO

Monte Carlo methods (direct and metamodel-based)

Outlier Analysis

Support Vector Machines (SVMs) for Statistical Classification

Taguchi Method

Curve similarity measures: Dynamic Time Warping (DTW), Partial Curve Mapping and Discrete Fréchet

Experimental Design: Space-filling, Full or Fractional Factorial, Latin Hypercube

Metamodels: Neural networks, Polynomials, Kriging and Support Vector Regression

Network-based job scheduling

LS-OPT is capable of performing optimization with multiple objectives, disciplines and load cases. It can account for uncertainties in the design or loading (stochastic analysis and optimization), and can also be used to optimize tolerances with a multi-level setup.

Graphical Post-Processing:

Result plots (Correlation Matrix, Scatter plots, Parallel Coordinate, Self-Organizing Maps, Time-history, Statistical)

Metamodel plots (Surface, 2D cross-sections, Accuracy, Global sensitivities, History sensitivities)

 Classification boundary plot

Pareto plots (Scatter plots, Parallel Coordinate, Self-Organizing Maps)

Stochastic Analysis (Statistical tools, Correlation, Stochastic Contribution)

Effect plot (Taguchi)

Optimization History

Tables with interactive features


LS-OPT GUI Defining Process Flow


Actual (LS-DYNA)       Classifier (blue outline)
Parametric Vehicle Intrusion Using a Classifier


GISSMO Failure Model Calibration Using DTW
Failure Model Calibration(LEFT)

Full Field Calibration (Digital Image Correlation)(RIGHT)

Material Parameter Identification

Go Back