darshanmakwana412 / structure_optimization

Optimizing Every Structure, Everywhere, all at Once

Optimizing Every Structure, Everywhere, all at Once

This project is currently work in progress

Ablation Studies

Initial Bridge	Minimizing Force	Minimizing Strain	Minimizing Strain Energy

Initial Tower	Minimizing Force	Minimizing Strain	Minimizing Strain Energy

Experiments

Branchistochrone Problem

The branchistochrone problem is the problem of finding a curve of shortest descent between two points, we assign a cost to each link as the total time it takes for a particle to travel it

Evolution of Topologies

Generation 10	Generation 20	Generation 30	Generation 40

Population 50	Population 60	Population 70	Population 80

Total Simulation Time: 41 Hours

To Do

• Create a 3D implementation of the original proposal
• Update Installtion instructions in the README
• Plot strain energy vs iterations
• Benchmark with different frameworks and architectures
• Implementation in C++
• Implementation in Javascript
• Create a project page for the project
• Implement a learning rate scheduler for greater convergence towards the optimal structure and decrease the wobbliness
• Maybe include the stiffness matrix into force calculations for more realistic structures
• Taking inspiration from NEAT(NeuroEvolution of Augmenting Topologies) implement an evolutionary approach for creating and linking nodes to achieve a robust and optimal structure

Experiments to Perform

• Create a tower and apply two forces on the top
• Add an additional constraint for volume minimization in the loss function, current hypothesis is that when opimized with gravity the structure will resemble a honeycomb
• Optimize bridges
• Create an infinite grid and simulate a hand punching in it

Useful References

• NEAT (https://ieeexplore.ieee.org/document/6790655)
• Concise summary of NEAT (https://nn.cs.utexas.edu/downloads/papers/stanley.ec02.pdf)

Installation

Clone this repository:

git clone https://github.com/darshanmakwana412/structure_optimization.git
cd structure_optimization

Using Docker (Recommended):

1. Build the Docker image:

   Run the following command to build the Docker image:

   docker build -t struct_optim .

2. Run the Docker container:

   Execute the following command to run the Docker container:

   docker run --gpus all struct_optim

Using Conda

Using Virtual Environment