Hydrostatic Curves of ship

[JoshTruman]

Extrapolating is not advisable if you need to perform any "delicate" calculations. I believe the only thing you could do is, given the curves of a similar ship, if the curves in the common areas are similar, perform an affine transformation of the model hull and deduce the curves in the missing draughts.

I guess it still a good topic for scientific research a comparison between model based calculation, and extrapolation by MLP's. Maybe in that case MLP's will be sufficient enough for that kind of work - no one is talking about prediction of whole hydrostatic - but only its extrapolation.

 

[TANSL]

Okay, go ahead and extrapolate. Using what criteria, linear extrapolation, parabolic extrapolation, .....? With all due respect, I think that would be absurd. But OK, if you don't mind, draw the missing curves however you like, and get a "nice" set of curves.

 

[JoshTruman]

Okay, go ahead and extrapolate. Using what criteria, linear extrapolation, parabolic extrapolation, .....? With all due respect, I think that would be absurd. But OK, if you don't mind, draw the missing curves however you like, and get a "nice" set of curves.

Thanks for advise - thats why wy have machine learning and MLP methods for ie forecasting based on known data (training set). I does not mean that this should be a method for primary use instead of integration of body lines it should be advisory in special condition like grounding or docking as mentioned before.

 

[TANSL]

I does not mean that this should be a method for primary use

Good luck, then, with your extrapolation, for secondary use.