William Luszczak  1:30 PM

This is the directory with the current PUEO antenna data files: https://github.com/PUEOCollaboration/pueoSim/tree/main/data/antennas. Each type of antenna will need several files:

• vv_0 and hh_0 describe the on axis v and h pol gain as a function of frequency. Gain is listed as a multiplicative factor (not in dB, if I'm remembering correctly)
• hv_0 and vh_0 describe the on axis cross polarization gain
• vv_az and hh_az describe the off-axis gain via multiplicative factors of the boresight gain. These are currently listed as a function of frequency for reference angles of 0,10,20,30,40,50,60,70,80, and 90 degrees, though we can always adjust the reference angles if needed.
• vv_el and hh_el are similar to above, but for gain as a function of elevation angle instead of azimuthal angle.

So for example, the total vpol gain in a particular frequency bin for a signal incident at azimuthal angle az and elevation angle el would be vv_0(f)*vv_az(f, az)*vv_el(f, el)

New

1:31

We can also potentially adjust this if it would be more convenient for the GENETIS people to output antenna information in a different format (or at different reference angles or whatever). The important thing is that the boresight h, v, and cross pol gains are all defined, as well as the off-axis response (as a function of azimuth and elevation)

Here are some ANITA/PUEO pictures that I received from Christian Mike that might be helpful.  I include his descriptions as well.

The one with the black crates in the background is the  ANITA 1 and ANITA 2 form factor and the one in the anechoic chamber is the ANITA 3 and ANITA 4 form factor. I dimensioned out an ANITA 3/4 antenna, in CAD space, to give you a better idea of the ridge profile. The ANITA 1/2 CAD models we have, I found, are inaccurate.

Attached are images of the PUEO antenna array geometry. One image annotates the approximate locations of the Vpol feed with the top antenna feed as the origin.

The following is a run summary of an AREA test using 100 individuals per generation, without the linear frequency dependence on the gain, with the frequency fixed (i.e. the same gain pattern is produced for all frequencies). This run was a preliminary test of the AREA algorithm with a larger NPop (100 individuals per generation) after seeing promising results in the previous test that only used 20 individuals per generation. The percentages of selection methods/operators used attempted to mimic the optimal percentages used for the PAEA Bicone Loop, but because the algorithms do not set these in the same manner, these are highly unlikely to be the best percentages to use. The results show a very quick plateau in fitness score with a loss in diversity of solutions after a low number of generations, potentially meaning that a poor breakdown of selection methods/operators was chosen. This result seems to underscore the need for a test loop-style optimization for AREA to determine the best breakdown of selection methods and operators to use for AREA.

Run details:

• Run Type
  • AREA
• Run Date
  • January 18, 2023
• Run Name
  • 20230118breynoldsrun_noLinearDependAREATest_38RCO_2RM_58TCO_2TM_150000NNU_2Seeds
• Why are we doing this run?
  • This test increased the number of individuals to 100 and attempted to mimic the percentages of selection methods and operators found to be optimal for the PAEA loop.
• What is different about this run from the last?
  • The previous test showed slow/minimal evolution after ~12 generations, and it was speculated that this was due to a small NPop of 20 individuals. This test uses an increased NPop of 100 individuals.
• Symmetric, asymmetric, linear, nonlinear (what order):
  • N/A
• Number of individuals (NPOP):
  • 100
• Number of neutrinos thrown in AraSim (NNT):
  • 300,000 Total (150,000 NNU x 2 Seeds)
• Operatiors used (% of each):
  • 96% Cross-Over, 4% Mutation
• Selection methods used (% of each):
  • 40% Roullette, 60% Tournament
• Are we using the database?
  • N/A

Results:

• Summary and comments on results
  • The fitness score quickly plaeaued and diversity of solutions was lost, potentially due to a unsuccessful choice of selection method and opterator percentages.
  • An important future step is to work with Ryan to create a test loop optimization for the AREA algorithm, as it is written differently in how it is passed selection method and operator percentages and requires its own optimization to find the optimal ratios of these. (Work on creating a test loop optimization for AREA is currently underway).

EDIT 4/10/23
Re-uploading gain pattern associated with most fit individual from this run

The following is a run summary of the AREA test without the linear frequency dependence on the gain, with the frequency fixed (i.e. the same gain pattern is produced for all frequencies). This run was a preliminary test of the AREA algorithm after fixing long-standing errors in the algorithm and finding that the initial constaints implemented did not work with the frequency dependence on gain. Results from this run did show a "proof-of-concept" that the AREA algorithm seems to evolve single frequencies in a promising way, however a low number of individuals may have stalled the evolution.

Run details:

• Run Type
  • AREA
• Run Date
  • December 26, 2022
• Run Name
  • 20221226breynoldsrun_noLinearDependAREATest_8RCO_8RM_2TCO_2TM_30000NNU_10Seeds
• Why are we doing this run?
  • We tested the AREA algorithm set to evolve at a single frequency in order to see if it would evolve a gain pattern at that frequency within the required constraints.
• What is different about this run from the last?
  • This run only evolves gain patterns at a single frequency, allowing us to investigate the evolved gain patterns and check that the required constraints on the power conservation are met, without yet fixing the constraints function to work with the linear frequency dependence relationship. This is also the longest test run since the AREA algorithm was fixed and gotten into working order.
• Symmetric, asymmetric, linear, nonlinear (what order):
  • N/A
• Number of individuals (NPOP):
  • 20
• Number of neutrinos thrown in AraSim (NNT):
  • 300,000 Total (30,000 NNU x 10 Seeds)
• Operatiors used (% of each):
  • 50% Cross-Over, 50% Mutation
• Selection methods used (% of each):
  • 80% Roullette, 20% Tournament
• Are we using the database?
  • N/A

Results:

• Summary and comments on results
  • The evolution appeared to be successful in that the most fit gain pattern found (attached) seems to make sense with our understanding of the underlying physics. Additionally, after inspecting the outputs, conservation of power was confirmed to be properly constrained. However, in the attached violin plot, the fitness score plateaued quickly and never evolved further, with the best individual coming in generation 12. This was speculated to be due to the small population used in this test (only 20 total individuals per generation).

I went through the current PUEO Loop  and documented everything that still needs to be accomplished before we can run.

Everything is compiled in this Google Doc which I will paste the current version of below.

Current Loop Overview:

Part A:

Runs GA, need to implement most recent GA (easy)

Part B:

B1:

Enters PUEO freq into simulation_PEC.xmacro then cats in PUEO simulationPECmacroskeletons.

Will  need to update simulationPECmacroskeletons with XF script to simulate the Hpol and Vpol sides of the antenna separately.

The bash portion should be working.

B2:

Currently almost the same as the ARA version. Creates the output.xmacro and moves the uan files. Will most likely need to change to get files in PUEO/icemc format, ready for input into the conversion script.

Part C:

Runs conversion script XFintoPUEO.py

The Bash portion is complete. We will need simple modifications to XFintoPUEO.py depending on XF outputs.

Part D:

D1:

Changes setup file, runs IceMCCall_Array.sh batch job

Need make this batch job to change IceMC / PUEO to read in the correct files 

(vv_0_{gen}_{i}, hh_0_{gen}_{i}) for the antenna i in generation gen.

I’ve commented where this should go. The rest of the job should be complete

The debug mode for D1 might not work, but it shouldn’t be necessary to run. 

So, the bash portion should be able to run.

D2:

Currently confirms runs by looking at the number of files in the AraSimConfirmed directory and looks for specific AraSim errors. 

This line will need to be changed to count the output files for icemc/pueo.

Change the error detection to just resubmit the job or something similar as we aren’t yet familiar with icemc errors during runtime. 

Part E:

Runs the PUEO fitness function. Unfortunately, icemc doesn’t output error bars, so we will have to remake any plots reliant on that. Veffective is also the same as fitness score.

Currently it should be working with the basic needed plots for gain patterns and veffective scores. But, we should add some more cool, automatic plots. 

Part F:

The plotting software seems to be specific to ARA output, needs PUEO versions.

ASYM LOOP:

Should be working with PUEO as a switch

IMPORTANT ISSUES:

Find out how to simulate polarization in XF:

Modify the current XF model, can be found in PAS1960/dylanwells1629/testproject.xf

The current model has all 4 sides of the antenna connected. However, the actual antenna is basically 2 antennas, one hpol and one vpol, electrically disconnected from each other. So, adjacent parts shouldn’t be electrically connected like they currently are in the model. We might have to simulate the two models independently. The goal for this portion is to figure out how to split up the antenna in the xmacro, and how to either simulate hpol and vpol separately or how to get xf to output hpol and vpol gain patterns.

XF Cross - Polarization, constraints --Alex

Make IceMC / PUEO read in out input gain patterns 

Will need to modify Batch_Jobs/IceMCCallArray.sh to change icemc to read in the current gain files before running. You can find the conversion script in Antenna_Performance_Metric/XFintoPUEO.py and the batch job will be run in Loop_Parts/Part_D 

We need to change icemc / pueo to read in the correct gain files before running it.

This will involve changing the anitaBuildTool/components/icemc/anita.cc ReadGains function starting on line 1413 to read in vv_0_{gen}_{num}, hh_0_{gen}_{num} , etc. (Do this with PueoSim) -- Dylan

I’m not sure if you will need to recompile these functions before running the icemc executable again, so find that out too. (enter pueoBuolder/ just run make) --Dylan

Make PUEO plotting software

All of the current plotting software is found within Part_E and Part_F. Sadly, icemc does not output errors for the veffectives (at least not that I could find, maybe you can find it), so much of the Ara plotting software will not work with PUEO outputs. The fitness scores csv will be the same format as ARA’s, and veffectives will be the same as fitness scores for PUEO. Figure out how to change the existing ARA python plotting scripts or make new ones.  --Bryan

Find out how to run PEUOsim and document outputs

Currently the loop is working with icemc, but we want to use PUEOsim for the future. So, document installing it, running it, and the outputs. (inputs are the same as icemc) -- Dylan 

We began a new run for a symmetric bicone antenna. The purpose of this is to satisfy the most recent comments from the reviewer on the paper. We are using the same GA as for the paper with the same parameters. I will update this post soon with more details about this run. There is a run_details file attached, but please wait for me to update this post with more specifics as the version of the GA being used is less connected to the details presented in that file.

The data can be found here: /fs/ess/PAS1960/BiconeEvolutionOSC/BiconeEvolution/current_antenna_evo_build/XF_Loop/Evolutionary_Loop/Run_Outputs/2023_02_20_Symmetric_Run .

AraSim Simulations for Top 5 Antennas from 12/2022 Run with 3 million simulated events

Simulated using Arasim version currently used in GENETIS Loop

/fs/ess/PAS1960/BiconeEvolutionOSC/AraSim/
 

AraSim ROOT files stored in

/fs/ess/PAS1960/ROOT_Files_Higher_Stats/AraSim_Loop_122022/Feb_2023_Crazy_Sides/

Antennas simulated

• Generation: 13, Individual: 84
• Generation: 18, Individual: 89
• Generation: 19, Individual: 96
• Generation: 29, Individual: 87
• Generation: 9, Individual: 50

Results

Conclusion

Antenna from Generation 13 Individual 84 performed the best with these stats

Attached is a pdf of physics plots generated for the 9_50, 13_84, 18_89, 19_96, and 29_87 crazy sides individuals along with another containing information on the amount of triggered events and effective volumes for each individual. The crazy sides individuals were simulated for 3 million events using the same Arasim version as is currently in the loop.

Input File Format for pueoSim (Also ICEMC)

Frequency Range: From 200 MHz to 1500 MHz incrementing by 10 MHz steps

There are 8 different files that are required for pueoSim. They are:

vv_0: Max Gain at each Frequency, Vertical Polarization

hh_0: Max Gain at each Frequency, Horizontal Polarization

vh_0: Max Gain at each Frequency, Vertical to Horizontal Polarization

hv_0: Max Gain at each Frequency, Horizontal to Vertical Polarization

vv_el: Gain at Theta Angles [5, 10, 20, 30, 45, 90] for each Frequency, Vertical Polarization

vv_az: Gain at Phi Angles [5, 10, 20, 30, 45, 90] for each Frequency, Vertical Polarization

hh_el: Gain at Theta Angles [5, 10, 20, 30, 45, 90] for each Frequency, Horizontal Polarization

hh_az: Gain at Phi Angles [5, 10, 20, 30, 45, 90] for each Frequency, Horizontal Polarization

Currently XF doesn't output all the information we need to create all these files. The only files able to be made from current XF outputs are vv_0, vv_el, and vv_az. Once we have the correct XF Outputs, it shouldn't be too much of a hassle to fix the current translation code.

Format for each of the files are the same. One column is Frequency (Hz) and the other is Gain. Attached are example files vv_0 and vv_el to visually see the format, though the frequency range is the current ARA range and not the final PUEO range as I am using ARA data to make these files. The files are named vv_0 (no .txt or .csv or any extension) and vv_el

Link to Google Doc with this information: https://docs.google.com/document/d/1iRUF6hIEyQfMK0LL21caRuHPgXYP30ZdkSkFv_-Y8R0/edit

Physics plots were created for 300K events from the higher statistic simulation results for the 9_50, 13_84, 18_89, 19_96, and 29_87 antennae. 

The simulation was inaccurate because a discrepancy exists between the higher simulation physics plots and the physics plots created for the 300K event results currently in the GENETIS Loop. This can be seen in the attached pdf file. 

The simulation may have been inaccurate due to incorrect gain files. 

The higher statistic simulations must be rerun for the 9_50, 13_84, 18_89, 19_96, and 29_87 antennae.

I've completed laying out the PCB for the 14-rung matching circuit. Attached are png files of two PCB designs. One uses LPS5050 inductors while the other uses LPS6235 inductors. The dimensions shown are in millimeters. 

Asym Straight Sides (from the paper - Generation 23, Individual 8)

(inner radius, length, opening angle in radians)

2.08711,89.924,0.0161734
0.30175,45.3616,0.0910478

individual found in 

/fs/project/PAS0654/BiconeEvolutionOSC/BiconeEvolution/current_antenna_evo_build/XF_Loop/Evolutionary_Loop/Run_Outputs/AraSim_Polarity_Fix_2021_03_19/AraSim_Polarity_Fix_2021_03_19.xf/Simulations/001108/Run0001/

Attached is a spreadsheet with the information on the parts we need for the N=14 matching circuit board.

https://docs.google.com/spreadsheets/d/1x8dX3tNE-WSHjH_slj_EH4XsHcAnCVC05XiRBFPtIUc/edit#gid=0

Here we will record where we obtained the ARA bicone antenna response files we use as our baselines. We want to record where we found them/who gave them to us and how they were generated (either through simulation or from actual tests).

Attached are the XF files for the top 5 individuals from the 2022_12_29 crazy sides run. In order from best to 5th best: 

1: Gen 29 Indiv 87

2: Gen 19 Indiv 96

3: Gen 13 Indiv 84

4: Gen 9 Indiv 50

5: Gen 18 Indiv 89

These can also be found in /fs/ess/PAS1960/BiconeEvolutionOSC/BiconeEvolution/current_antenna_evo_build/XF_Loop/Evolutionary_Loop/Run_Outputs/2022_12_29/2022_12_29.xf in the corresponding directories. 

We began a new run of the curved sided GA. Previously, we conducted a run just like this in July of 2022 (titled 2022_07_15_Latest_Greatest). In that run, we made substantial updates to the loop, including using the latest version of AraSim. However, we discovered after the run (which had a very flat fitness score growth curve) that we had not modified AraSim to take in gainfiles appropriately (so it was just evaluating the ARA bicone repeatedly). We've fixed this issue and a few more (see the github repository for the issues that exist and have been resolved) and are redoing this run. The run details will look much the same as that one and are visible in the attached file. Here is the location of the directory where the data for this run will be stored: /fs/ess/PAS1960/BiconeEvolutionOSC/BiconeEvolution/current_antenna_evo_build/XF_Loop/Evolutionary_Loop/Run_Outputs/2022_12_29

For a list of changes made in this run compared to previous ones, see this ELOG post on the previous run: http://radiorm.physics.ohio-state.edu/elog/GENETIS/186

Variables of Evolved Antennas

Wall variables:
S -- length of bottom of the walls
m -- slope of the walls (currently set to 1)
H -- height of walls
x0, y0, z0 -- 3D coordinates of the bottom of the ridge
xf, yf, zf -- 3D coordinates of the top of the ridge
Beta -- curvature of the ridge

Antenna Walls

S is half the side length of the bottom wall
m is the slope of the outer wall
H is the max height of the outer wall

Current Constraints:

S < 50cm
H < 50cm
m = 1

Antenna Ridges

x_0, y_0, z_0 are the innitial points of the inner most part of the ridge
x_f, y_f, z_f are the final points of the inner most part of the ridge
tau is the parametric time range beta is the slope of the curve of the ridge

Current Constraints:

tau = 0.26
x_f = S
0 < x_0 <= x_f
0 < y_0 <= x_0
0 < y_f < z_f
z_0 = 0
0 < z_f <= H
(4/30) * z_f < beta < 7 * z_f (for z_f in meters, designs will 'compile' in xfdtd above this upper limit, but the curve is functionally a line for all values greater)

Here are the changes that have been made in this (and the preceding) run relative to what we have been doing before.

1. The newest version of AraSim has been implemented.
   1. This was done because AraSim has gone through significant changes since 2019 that make it more accurate and makes it run faster.
2. The minimum length has been decreased to 10 cm per side.
   1. This was done because we previously tested shorter lengths for a response to the reviewer when we submitting the paper to Phys Rev D. Previously, we thought that going below 37.5 cm would lead to unreliable results, but we are reassured that that is not the case now.
3. The GA is recording the parents and operators used to generate individuals.
   1. This should allow us to look back and make a "history" of the evolution by seeing where genes come from in each generation.
4. The number of individuals has been increased to 100 per generation.
   1. Ryan demonstrated that a higher number of individuals does lead to a substantial improvement in the speed at which the evolution converges. 
      1. We previously thought that it did not make a big difference. However, that was done by measuring the average maximum fitness score. If we measure how many generations it takes for the evolution to reach a benchmark score, we find that more individuals helps significantly.
5. The number of neutrinos per individual has been increased to 600k.
   1. This was done to make the fitness scores more precise. At 300k neutrinos, we expect an error of around 0.2. This will cut things down closer to 0.1, which may help us decrease the number of outlier measurements without increasing run time too much because of the AraSim speed up.

This run may take longer than previous runs due to the increased number of individuals. There may need to be modifications made to resolve or work around AraSim errors that delay the loop (due to the auto-resubmit function). Find the details of the run attached.

This is the plot for the run type that had the best performance in the most recent optimization run that was completed. The most optimized runs from these collections are decided by which run types achieve a chi-squared value of 0.1 in the fewest generations (correlating to about a 0.9 fitness score). However, the genetic algorithm does not use Chi-squared directly, instead trying to maximize the fitness score (which uses a chi-squared in the denominator) that simulates the error we see in arasim. This fitness score is usually maximized at 1.0 but the simulated error can push the score slightly beyond this. This test was conducted with a simulated error of 0.1. This optimization holds the selection methods steady at 2 tournament 2 roulette and 6 rank selected individuals per 10 selection events. The changing parameter values are in the amounts of Reproduction, Crossover, Immigration, and mutation (rate and width) that the GA performs each generation. Each generation uses 100 individuals. In the graph below, the dotted lines are the minimum Chi^2 values at each generation, solid lines are the maximum fitness score. Each color represents a different run using the same parameters and dotted and solid lines sharing the same color represent the same run. The prevailing trend in the tests leading up to this test has been that the most optimized runs have low amounts of reproduction, high amounts of crossover, and they have been pushing the boundaries of the mutation rate and gaussian width of those mutations and this run follows that trend. The ranges of values we tested over were 0-12R, 88-100C, 15-25M_Rate, and 3-7 sigma, up from the previous test. This run took an average of 16.8 +/- 11.9 generations to reach the 0.1 chi-squared benchmark, the next closest run took 20.3 +/- 10.9 generations to reach the same point, while the most optimized run from an earlier test came in as the 10th best run taking 21.9 +/- 14.4 generations. The graph shows the behavior of this by showing large improvements in early generations that tend to quickly plateau around the maximum value before oscillating around the peak it reaches. Oscillations could be caused by either small mutations around our answer and displacement from the simulated error. The Chi-squared values show a much smoother behavior with smaller oscillations at the end of the run, which helps demonstrate the impact of the error on the behavior.

Here is our current assumption of the antenna angles needed for the icemc inputs.