Hello AWR users,
I am trying to optimize a RC-circuit (only passive elements) and created a measurment (voltage gain) and set optimizer goals for this measurement, but the goals are not appearing on the graph. I noticed that some of the goals (I have 7 goals in total) are visible on the graph when they are disabled. These goals dissappear when I enable them again. Also my goals' costs do not converge to 0 during otimization using different methods.
Could you please advise how I can add my goals on the graph?
I have acedemic licence for AWR MWO version 11.04r build 7062 64-bit.
A couple of things to check. Graphs will scale their axes by the simulation data by default so you goal may be off of the visible scale of your simulation data. You can fix the axis scaling in the graph properties or change your goal to be in the same range as your simulation data to see if that is causing the issue. There is another option that might be causing issues. Go to Options > Environment Options on the Project tab. There is an option called "Diabled goals draw on graph" that effects how goals that are disabled draw.
If these don't solve the problem, please attached the project so we can take a look.
thank you for your response!
1. My goals are scaled correctly (within the range of the measurement). So this cannot be the reason.
2. I found the way around how to visualise my goals. There are a measurement on my graph and the imported curve (two columns plotted from the "data files" using PlotCol measurement). So if I remove the PlotCol measurement the goals appear on the graph!
3. Option "Diabled goals draw on graph" helps not to see the goals when they are disabled. But it does not help to see the goals when there are more than one measurement.
Visulaisation of goals is less important than the actual results of optimization based on these golas.
I am trying to model experimental data for voltage gain (imported file) with RC-filter cascade. Is there a way to take into account the whole frequency spectra? My goals are set based on the imported data set - I just picked up several points to discribe the curve. It is a best way I know so far. I have tried different methods but I cannot fit the data with my model circuit.
I would very much appriciate if you could help me to improve my optimization.
I have uploaded the project here:
look forward to your comments
Thanks for the project, that made this very clear. I have attached my response here:
For the first issue, the problem is the format of your data. There are no units in this data and so goals and comparing with other data can be difficult. You can tell it the units, but that is harder. I have shown a good technique to do your modeling. Essentially, we calculdate a delta between the two reponses. Then you can optimize that delta to zero. To do this, I had to do a few thigs.
1. I did way fewer simulation points, 40 instead of 4000. Will be much faster. You can change this if you want.
2. I added output equations to calculate the delta. Since you measured and simulated data have different frequency points, we have to interpolate the data to be on the same range as the simulation to do the delta. See the "Output Equations" document to see how I did this.
3. You would then just need one optimization goal to get the "delta" measurement to go to 0.
I think you will find this a much cleaner way to do what you need.
Many thanks for the prompt feedback! Now I got the point about the goals visualisation.
Concenrning the optimization it sounds very promising - I was sure there should be a better way to do it.
Before I can follow your steps, could you please check the access settings for your uploaded file?
It seems that I am "unauthorized". The full message is the following: "It appears you're not allowed to view what you requested. You might contact your administrator if you think this is a mistake."
I think the "administrator" the message is talking about is your university's IT system administrator. I don't think that message comes from AWR design environment.
Sorry about that. Please try the document again as I now made the permission for anyone to view.
sorry for late reply. Many thanks for the help! The approach you have shown is great, I've learned quite a few things