-2

I have reproduced the work of Grady et alenter image description here, with the wind in single direction and multiple direction. I have not obtained similar layout though same power(kW/year)

I was wondering if using the case objective function in Genetic Agorithm in Matlab will produce same layout?

Or will it produce different layouts?

Plz advice

fredd
  • 39
  • 1
  • 2
  • 8
  • Are we supposed to go and read about the problem and come back and try to understand the question from the 'et al' things that you have written? Please explain your problem properly so that people can help you. – Hazem Apr 21 '18 at 09:00
  • I apologize @Hazem I have edited my question. – fredd Apr 21 '18 at 09:09
  • 1
    For this to be an answerable question you need to add: 1) an explanation of what Grady et al's work is about, 2) the code you used to get the results you don't like, 3) a description of why these results are different from what you expect. See [how to ask](https://stackoverflow.com/help/how-to-ask). – Cris Luengo Apr 21 '18 at 18:17
  • @CrisLuengo, my code is 5 files long. I though someone from Genetic algorithm can anwer this. ie my question is whether we should all obtain same exact layout when using GA or not when reproducing other's work. Grady et al have worked in a wind farm – fredd Apr 22 '18 at 04:20
  • @CrisLuengo is absolutely right. I would also add that you have not even provided any pointers to the actual paper/book/whatever the original results are. Googling "Grady et al" really does not help. However, even if you provided a proper link, I would still not consider it to be a good question. You should at least summarize what the original work was about and what methods were used, how exactly did you reproduce it, what are some potential differences and in what regard do you consider your results to be different. GAs are stochastic algorithms after all, there are no two equal results. – zegkljan Apr 22 '18 at 22:53

1 Answers1

0

As were pointed in comments under your question we greatly lack the context and details of the problem you're solving and issues you're doubting about. Still...

I would highlight that genetic algorithms in general involve a lot of randomness, so exact reproduction of any previous results is practically impossible. You can expect your results to expose similar characteristics, though; but that's what you achieved.

Of course, the difference you're facing can still be important, but this is impossible to say after such a little introduction to the problem.

werediver
  • 4,667
  • 1
  • 29
  • 49
  • the fact that you told me that '...in general involve a lot of randomness, so exact reproduction of any previous results is practically impossible. ' is enough. You have replied to my question.. Thanks tons. – fredd Apr 23 '18 at 06:56
  • 2
    @fredd Don't get too frustrated with the negative feedback you had with this question. People on SO really like well shaped questions, but as soon as you managed to prepare one you get good chances to find help here! Also consider joining other [Stack Exchange sites](https://stackexchange.com/sites). – werediver Apr 23 '18 at 10:34
  • Again thanks heaps. .Actually I was banging my head against the wall with this. I though i would obtain same layout as other authors. – fredd Apr 27 '18 at 10:16