Timeline for Frequency dependent impedance as an electrical circuit
Current License: CC BY-SA 4.0
11 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 9, 2020 at 10:50 | comment | added | Wallflower | Understood, thank you so much! | |
| Sep 9, 2020 at 10:49 | comment | added | Neil_UK | @Wallflower yet another paragraph, this is part of 'getting to know your problem'. If it's well behaved, lucky you, you'll get away with not needing to do too much work, it will solve itself. That large frequency range concerns me, the sensitivity to each of the component values will change with the frequency | |
| Sep 9, 2020 at 10:46 | history | edited | Neil_UK | CC BY-SA 4.0 |
added 496 characters in body
|
| Sep 9, 2020 at 9:28 | history | edited | Neil_UK | CC BY-SA 4.0 |
added 93 characters in body
|
| Sep 9, 2020 at 8:54 | comment | added | Wallflower | @NeilUK how can I share with you my scatter plot? I have a 500 points vector in my scatter plot which I obtained from a FEM simulation over the following frequency range: f = 1e6:1e6:4.99e8. The scatter plot is of real values only as the function is a frequency dependent resistor so no phase is defined. Based on the paragraph you've added, my objective function should be: Z_FEM(f)-Z_circuit(f) which should be minimized using NM method. Ok I will try this approach, thank you @Neil_UK! | |
| Sep 9, 2020 at 8:05 | comment | added | Neil_UK | @Wallflower I've added another paragraph. I'm not going to give detailed advice without seeing the data. I know 'get to know your problem' seems a bit lame, but machine optimisation is fraught with underspecifying its constraints, and then it goes off and meets them in an unexpected way. So you need to know at least roughly what the landscape of solutions is going to look like. | |
| Sep 9, 2020 at 8:02 | history | edited | Neil_UK | CC BY-SA 4.0 |
added 1303 characters in body
|
| Sep 8, 2020 at 8:39 | comment | added | Wallflower | Neil_UK, thanks for the reply. Yes, I already have the analytical expression of the equivalent impedance of the target circuit shown in the picture above. Concerning the first attempt, I cannot do so since Z(f) is a scatter plot and not an analytical function. The second alternative is kind of unfit for my application, since I have 56 Z(f) scatter plots. Doing it by hand is surely not the best option. I am more interested in fitting scripts/algorithms that will enable me to custom my target fitting function and take an an input the scatter plot so as to automate as much as possible my work. | |
| Sep 8, 2020 at 8:31 | history | edited | Neil_UK | CC BY-SA 4.0 |
added 6 characters in body
|
| Sep 8, 2020 at 8:26 | vote | accept | Wallflower | ||
| Sep 8, 2020 at 8:24 | history | answered | Neil_UK | CC BY-SA 4.0 |