| ... | ... | @@ -6,13 +6,19 @@ The PSI Ring parameters are not well-known. The goal of this study is to find go |
|
|
|
The OPAL probe code has been extended to find the radial turn locations in the same way as in the measurement, see figure below.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
RRI2 probe _simulation_ histogram (for 5000 particles) with found peak locations.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
RRI2 probe _measurement_ with peak locations.
|
|
|
|
|
|
|
|
Test 1
|
|
|
|
[Up to peak 5](#initial-test)
|
|
|
|
[Improvements](#improvements)
|
|
|
|
|
|
|
|
Initial Test
|
|
|
|
=======
|
|
|
|
|
|
|
|
First trial of optimizing three objectives:
|
|
|
|
|
|
|
|
* difference between simulation and measurement of peak 1
|
| ... | ... | @@ -60,13 +66,81 @@ Sum of all differences (simulation - measurement) of objective 3 per generation. |
|
|
|
|
|
|
|
We see a nice decreasing curve for peak 1. Also the curve for peak 2 is mainly decreasing. However, the third objective is also increasing over the generations.
|
|
|
|
|
|
|
|
Results for one of the optimised results:
|
|
|
|
Results for one of the "best" (defined as the minimum of the sum of objectives) seeds:
|
|
|
|
|
|
|
|
```
|
|
|
|
%ID %dpeak1 %dpeak2 %dpeak3_5 DVAR: %benergy %phiinit %prinit %rinit
|
|
|
|
25 42 4.500000e+00 3.000000e-01 5.174510e+00 7.120380e-02 1.098186e+02 -1.513638e-02 2.030688e+03
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
Probe measurement, simulation and measurement (scaled)
|
|
|
|
|
|
|
|
|
|
|
|
Radial peak differences between simulation and measurement versus peak number.
|
|
|
|
|
|
|
|
Test 2
|
|
|
|
======
|
|
|
|
|
|
|
|
After the first test a number of changes were applied:
|
|
|
|
|
|
|
|
* The RRI2 measurement was changed to RRI2-0065Y08. This is a 300uA Strahlentwicklung measurement by Jianjun.
|
|
|
|
* The number of generations was increased to 10000.
|
|
|
|
* The cavity voltages were adapted to the measurement setting (looked up with `analyze`).
|
|
|
|
* B-Field of trim coils 1 (turns 1-4) and 2 (turns 1-7) were added to the design variables.
|
|
|
|
* Objectives: dpeak1, dpeak2, dpeak3_4, dpeak5_7
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
One of the best seeds was again selected:
|
|
|
|
|
|
|
|
```
|
|
|
|
%ID %dpeak1 %dpeak2 %dpeak3_4 %dpeak5_7 DVAR: %benergy %phiinit %prinit %rinit %tc01mb %tc02mb
|
|
|
|
51 2.720000e+00 3.200000e-01 1.444022e+00 2.854766e+00 7.168652e-02 1.102582e+02 -1.438639e-02 2.040623e+03 -2.725981e-02 2.563253e-02
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Probe measurement, simulation and measurement (scaled)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Radial peak differences between simulation and measurement versus peak number.
|
|
|
|
|
|
|
|
Test 3
|
|
|
|
====
|
|
|
|
|
|
|
|
* Add trim coil 3 (turns 1-13) to design variables
|
|
|
|
* Reduce range of design variables based on previous solutions
|
|
|
|
* Increase the initial population from 62 to 126. Due to the increased number of design variables the phase space is significantly increased.
|
|
|
|
* Reduce number of generations to 1000
|
|
|
|
* Objectives: dpeak1, dpeak2, dpeak3_4, dpeak5_7, dpeak8_13
|
|
|
|
|
|
|
|
```
|
|
|
|
%ID %dpeak1 %dpeak2 %dpeak3_4 %dpeak5_7 %dpeak8_13 DVAR: %benergy %phiinit %prinit %rinit %tc01mb %tc02mb %tc03mb
|
|
|
|
143 1.260000e+00 3.400000e-01 1.380507e+00 6.554727e-01 1.314424e+00 7.158339e-02 1.096083e+02 -1.760224e-02 2.043670e+03 -8.025304e-04 -2.966774e-03 -1.287497e-04
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Video of the evolution of the pareto-front.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Zoomed Pareto-front. Some more generations might still be useful.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Probe measurement, simulation and measurement (scaled)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Radial peak differences between simulation and measurement versus peak number.
|
|
|
|
|
|
|
|
Test 4
|
|
|
|
====
|
|
|
|
|
|
|
|
The match is limited by the probe histogram bin size. This is expected to match better by:
|
|
|
|
|
|
|
|
* Reduce bin size of simulation probe.
|
|
|
|
* Increase number of particles |
|
|
\ No newline at end of file |
