I am undertaking experimental study on modifying asphalt binder using different percentages of different materials to enhance asphalt resistance to cracking and deformation and was wondering if there is a way to optimize this whole process to find the best percent that would provide the best properties. Keeping in mind that, for the most part, cracking and deformation are inversely related.
Not sure of your question. Do you want an experimental study or an optimization? Experimentally, perform a design of experiments (a la Taguchi for instance), varying each parameter at a low, medium and high level, running the experiments, and generating a way to assess the effects of each parameter on the process. If you can model the process, you can tie it to an optimization, and you would have to be able to compute some metric which represents what you want to obtain.
An alternative is to use the results of the experiment explained above to generate many results, then train a Neural network to reproduce the results, and use that Neural Network as your function, tying it to an optimizer to find the appropriate parameters...
Good luck
Thanks George! I was only asking about the optimization part. A simple procedure would be highly appreciated.
I don't know what you have in terms of function. It depends on that. Is it a continuous function, is it experimental data?
Dear Mohammad Khasawneh,
To my best knowledge, differential evolution, particle swarm optimization and cuckoo search have such characteristics. Among the aforementioned algorithms, however, DE is of interest.
http://www.particleswarm.info/Programs.html
http://www1.icsi.berkeley.edu/~storn/code.html
https://www.mathworks.com/matlabcentral/fileexchange/29809-cuckoo-search--cs--algorithm
Thanks Hojjat! Is it possible to use optimization approaches of excel solver?
The attached file implemented the stat-of-the-art optimization algorithms in excel and you can use it.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators