New application
log in

Advanced search

Message boards : News : New application

Author Message
Vadimas Starikovicius
Project administrator
Project developer
Project scientist
Send message
Joined: 3 Jan 12
Posts: 80
Credit: 273,126
RAC: 706
Message 226 - Posted: 19 Jul 2017, 17:25:31 UTC
Last modified: 19 Jul 2017, 17:26:31 UTC

Dear VGTU project@Home users,

We have decided to stop the computations with the current application for dynamic visual cryptography based on non-linear oscillations. Some results have been obtained and published. However, new ideas are needed to move forward.

In the meantime, we have developed a new application in collaboration with our colleagues from the VGTU Department of Information Technologies. New application is solving the grillage optimization problem with parallel global optimization algorithms. More detailed description is here.

We thank you for your active participation and hope that you will find the new application also worthy of your computer’s time.

We have tried our best to test it. However, some undiscovered issues may appear. So, we will start with small portions of work.

With best regards,
The Project Team.

__Dutch__
Send message
Joined: 29 Jun 17
Posts: 1
Credit: 41,048,795
RAC: 408,465
Message 237 - Posted: 21 Sep 2017, 10:47:27 UTC - in response to Message 226.

Hi Vadimas,

Is there any chance you could please provide us with an update of the current work? I understand that we are contributing to an optimisation problem, so is there an ETA for exhaustion of the current project's WUs?

Any chance you could give us a sneak peak preview of what we can expect to be working on once the grillage optimisation is complete?

Thank you very much for your time!

Best wishes,
Dutch

Vadimas Starikovicius
Project administrator
Project developer
Project scientist
Send message
Joined: 3 Jan 12
Posts: 80
Credit: 273,126
RAC: 706
Message 239 - Posted: 22 Sep 2017, 12:50:38 UTC - in response to Message 237.

OK, here is more information. Current problem of grillage optimization cannot be solved with some final number of WUs and closed. And the goal is not to solve the problem. We use it as so called “benchmark problem” for development and testing of parallel global optimization algorithms. Currently, we solve the problem with 55 unknowns using parallel genetic algorithm. We are testing the performance of algorithm and its modifications with different parameters.

After 2-3 months, we plan to implement and start testing algorithm based on Simulated annealing (global optimization technique). Using the results obtained on BOINC and HPC clusters, we are planning to prepare 1 or 2 publications.

So, this is the plan for the next 6 month.

Afterwards, depending on the results, we may continue with the current problem or move to some new problem, as has happened already several times in our project.

Regards

Message boards : News : New application


Main page · Your account · Message boards


Copyright © 2017 VGTU