de
en
Schliessen
Detailsuche
Bibliotheken
Projekt
Impressum
Datenschutz
de
en
Schliessen
Impressum
Datenschutz
zum Inhalt
Detailsuche
Schnellsuche:
OK
Ergebnisliste
Titel
Titel
Inhalt
Inhalt
Seite
Seite
Im Dokument suchen
Process for extraction of knowledge from crash simulations by means of dimensionality reduction and rule mining / submitted by: Constantin Diez. Wuppertal, February 16, 2019
Inhalt
Acknowledgements
Notation
Introduction
Vehicle Crash Simulation
Variational Studies
The Key Questions of Structural Behavior Analysis
Side Challenges
Goal of this Thesis
Related Work
Data Analytics Process Flow
Data Generation
Searching and Detecting Effects
Analyzing Simulation Responses
Deformation Mode Segmentation
Relating Cause and Effect by means of Inference
Mathematical Formulations
Deformation Class Segmentation
Geometry-Based Dimensionality Reduction
Geometric Simplification
Results Projection and Smoothing
Computation of Similarity
Clustering Similar Results
Visualization with Low-Dimensional Embeddings
Clustering
Decision Tree Learning
Decision Tree Basics
Random Forests
Rule Mining for Knowledge Extraction from Simulation Data
Example Data
Target Definition
Rule Extraction
Automatic Rule Filtering
Example for Rule Mining
Making Rules Reliable
Examples
Buckling analysis of a cantilever plate
Model Description
Results of the Buckling Analysis
Analysis of a Car Crash
Model Description
Results of the Structural Components
Analysis of the Bumper
Analysis of the crossbeam
Analysis of the left rail
Analysis of the Right Rail
Analysis of Responses
Summary and Conclusion
Discussion of Dimensionality Reduction
Geometric Simplification
Projection and Smoothing
Discussion of Rule Mining
Insights of Decision Tree Learning
Limits, Flaws and Unknown Potential
Description of the Longitudinal Rail
Bibliography