Identifying topological prototypes using deep point cloud autoencoder networks

Keyphrases

• Autoencoder Network 100%
• Topology Optimization 100%
• Point Cloud Autoencoder 100%
• Design Space 66%
• Geometric Features 66%
• Performance Objectives 66%
• Point Cloud Representation 66%
• Stiffness 33%
• Challenging Tasks 33%
• Design Process 33%
• Generation Method 33%
• Performance Features 33%
• Design Type 33%
• Engineering Design 33%
• Mass Fraction 33%
• Group Design 33%
• Topological Methods 33%
• Feasible Design 33%
• Design Ideas 33%
• Truss-like 33%
• Deep Autoencoder Network 33%
• Design Generation 33%
• Latent Code 33%
• Cluster Topology 33%

Engineering

• Point Cloud 100%
• Autoencoder 100%
• Optimization Method 66%
• Geometric Feature 66%
• Design Space 66%
• Design Process 33%
• Sampling Point 33%
• Initial Stage 33%
• Extracted Feature 33%
• Design Engineering 33%
• Mass Fraction 33%
• Feasible Design 33%
• Type Design 33%
• Boundary Condition 33%