Page 1 of 1

HARP_opt help on hydrokinetic turbine

Posted: Wed Mar 01, 2017 6:36 pm
by Yao.Wang
Hi,

I am Yao Wang, and running the HARP_opt to perform an optimization for a hydro-kinetic turbine. I successfully installed it as a stand alone software which could run without MATLAB. The user interface and the inputs are in the attachments. But it takes too long to find the feasible initial populations, it found only 1 out of 980 creations. Could anyone have a look into the inputs that I have used and see if there is anything that seems unreasonable or improper?

Thanks for the help,
Yao

Re: HARP_opt help on hydrokinetic turbine

Posted: Wed Mar 08, 2017 9:37 am
by Danny.Sale
hello Yao Wang,

I think the difficulty is related to cavitation. First attempt to disable the cavitation option, and hopefully that works. But there are several ways to change the design so cavitation is less likely; there is some other discussion about cavitation on the forums here: viewtopic.php?f=26&t=875&p=3630&hilit=cavitation#p3630

hope that helps,
Danny

Re: HARP_opt help on hydrokinetic turbine

Posted: Wed Mar 08, 2017 1:53 pm
by Yao.Wang
Hi Danny,

Thanks for your reply! I would try it and see if I could get any improvements.

Best,
Yao

Re: HARP_opt help on hydrokinetic turbine

Posted: Thu Mar 09, 2017 1:50 pm
by Yao.Wang
Hi Danny,

I have tried your suggestion of disabling the cavitation option. It did work! Thanks for your comments.

I met a new problem which might relate with the number of PopSize, or the NumGen. I tried several combinations, for example, (1) PopSize -50, NumGen - 10, (2) PopSize -50, NumGen - 20, ...etc

However, for some cases, not all of them, the iterations went through but the results couldn't be output successfully, the error says the "CAT arguments dimension are not consistent", but what does this mean?

I installed the HARP_Opt as a stand alone software which doesn't require the MATLAB installation, but I still have access to MATLAB and able to check the details in the source code. But I still need some help on it.

In the code " Post_Process" Line 153, it is "AEP = get_AEP([V PWR],pUvars);", are the CAT arguments some variable in this command? If so, why would the dimension mismatch issue happen in some cases?


Thanks!

Best,
Yao

Re: HARP_opt help on hydrokinetic turbine

Posted: Thu Mar 09, 2017 1:54 pm
by Yao.Wang
Also, for some of the cases, negative power and power coefficient occurs in the output excel file, is any suggestion on this phenomena?

Thanks,
Yao

Re: HARP_opt help on hydrokinetic turbine

Posted: Mon Mar 13, 2017 3:48 pm
by Danny.Sale
hi Yao,

I would check the content of the pUvars variable. This is related to the velocity probability distribution, maybe there is some mistake such that size of pUvars is incorrect. Or the V and PWR dimensions are not consistent for some reason. It's good you are running in Matlab, you can use the debugging stop points to further diagnose.

I would check on the Matlab toolbox documentation how to best set the GA population sizes, I cannot recall exactly.

Regarding negative power, that just means the turbine will not product power at those speeds. It is the raw result of the BEM analysis.

best luck,
Danny