Learning Forum

[atfkerem76]

Hello. I'm trying to optimize the thickness of a single layer structure in order to minimize its reflection, like in this video:

 

Even though I followed up the same steps, every time I try to perform an optimization I got this error:

Here is the desing and component tree of my simulation:

"R_" is the monitor and I'm trying to extract value of "T". To do that I created a single layer structure inside of structure group and added a parameter which is "thickness1".

Then I added this script:

And lastly, I made an optimization configuration:

After finishing the first generation of optimization it gives me this error, "The results for R_ are incorrect size.":

I'm trying to fix this for 2 weeks and still no progress. I tried to follow up same steps in video I shared but have no idea about the problem. Any suggestions would be great. 

[Amrita Pati]

Hello,

What is the size of R_, which has been extracted from the monitor T? You can use the size script command to do this. Are you recording this data at multiple frequency points? The in-built PSO algorithm in FDTD can only optimize a single-valued figure of merit. This issue might arise if R_ is an array/matrix. Please let me know if this doesn't work. Thank you!

Regards,
Amrita

[atfkerem76]

 

Hello again. I was recording this data at multiple frequency points (1500). Appearently this was the problem. I tried this with a single frequency point and now it works. Many thanks. 

I have some additional questions. They might be found on documentation, but I think answering these questions will be helpful for me and anyone who is reading this post.

 

1. Normally, the frequency points was set to 1500, so is it possible to see lower sensivity when we reduce this to 1? I mean reducing the frequency points means lower sensitivity?
2. If so, are there any other algortihms that can work with multiple frequency points? (It doesn’t have to be in-built). 
3. Right now I can optimize "T" value from a reflection monitor and transmission monitor (I placed reflection monitor front of the structure and placed transmission monitor back to the structure that could be seen in previous images.). I was wondering if it is possible to write a figure of merit script that can analyze "absorption" using this T values from reflection and transmission monitor. Normally I was doing it like this: 

   Trans = getresult("monitor_Trans","T");

   Ref = getresult("monitor_Ref","T");

   Transit = Trans.T;

   Reflect = Ref.T;

   absorb=1 - abs(Transit) - abs(Reflect);

   lambda=Ref.lambda;

   #plot(lambda,absorb);

   plot(lambda,absorb,abs(Transit),abs(Reflect));

   Using this script I was plotting "absorption" value using the extracted "T" datas from monitor_Ref (reflection monitor) and monitor_Trans (transmission monitor). Using something similar to this, is it possible to optimize absorption value? If so, how would the fom script look like?