These weeks' work has been devoted to the following activities:
- ECOS 2015 paper extension
- Funding application for H2020 application on LNG for shipping
- MSc project for control optimization of hybrid tug
- Paper on application of machine learning to optimization of maintenance for ships
- Abstract for SCC conference
ECOS 2015 paper extension
We are almost there! Now, almost all required calculations are performed, a large part of the code is appropriately debugged, and we can start getting some useful data out.
Hereafter, some more updates
- Finally, the calculation of the heat demand has been implemented. This time, I DO NOT use the information about the boilers' fuel consumption to calculate the demand, but I rather use that information as validation. The Result shows still potential for improvement but...we finally are getting something
- Finally, the calculation of electric power demand was also implemented. The "major" effort (since electric power generation is measured onboard, so the determination of the total demand was trivial) was to identify the consumption of the thrusters and the HVAC. Some assumptions had to be made, and we will discuss them internally before publishing
- I also implemented the analysis/modeling code for the heat recovery water system. This approach, based on the epsNTU method, allows for a much nicer estimation of the contribution of the HT cooling water to the onboard heat demand
- Finally, the steam systems were also implemented
So, lots of things are moving. The nice thing is: the type of system analysis that I implemented made it MUCH EASIER to implement changes: since I define balance equations and connections, the algorithm is (at least partially) able to automatically adapt to changes in the system structure
Funding application for H2020 application on LNG for shipping
On friday the 2017-07-07 we had the new project meeting. Nothing much interesting to say, but things are moving forward
MSc project for control optimization of hybrid
The contact previously representing Damen is leaving the company. But no worries, we have two more! One is Erik-Jan Boonen, someone I knew already and with whom I had been able to collaborate with ease. The other one is Peter Rampen, who I had met once before and who was very keen on keeping the collaboration alive.
The project idea moved from looking at a tug to a ferry (no problem) and aims at looking a bit into adaptive learning control (maybe a bit tougher). We see how it goes, but they seem motivated!
Paper on application of machine learning to optimization of maintenance for ships
Some big updates here as well. We contacted Par Brandholm at Laurin for some feedback and more data, and he responded very promptly! We (probably) got access to the Laurin/Marorka database again, and it seems (after some discussion) that we might have made our model work decently. And we can apply it to one more ship.
We are now working on comparing the white-box models (the real one from Andrea and the "fake" one I wrote according to the instructions provided by the ISO. Let's see what it gives, the process is ongoing!
Absract to Shipping in Changing Climate Conference
I submitted an abstract for participating to the "Shipping in Changing Climates" conference in early September (4-6) 2017. The presentation will deal with a very standard application of process integration to a ship case.
The main question now is: will I make it in time to prepare something?