New implementation of ARH (Active Radar Homing logic for Fox3)?
-
Entire new seeker aquisition parameters.
It’s now based on a Gaussian distribution errors , with less errors but more restricted aquisition window.
In practice against manoeuvering targets, same behavior (read PG ) than before (but with less trajectory deviation) but more reliable against non manoeuvering targets (in particular in TWS for instance )
-
@Stevie said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@suhkoi69 - insider trading…
Bottom line is that it was more “correct” before it was “fixed”, and behaved more as I would expect.
The previous modeling ( < 4.37) was a giant hack with absolutly no aquisition phase from the AIM120 seeker and with a PG of 100%
U3 improve the PG against non manoeuvering target, which was the major flaw of the previous (4.37) implementation.
you might take some time to test U3 before commenting
-
@Mav-jp said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@Stevie said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@suhkoi69 - insider trading…
Bottom line is that it was more “correct” before it was “fixed”, and behaved more as I would expect.
no it was not
The previous modeling ( < 4.37) was a giant hack with absolutly no aquisition phase from the AIM120 seeker and with a PG of 100%
U3 improve the PG against non manoeuvering target, which was the major flaw of the previous implementation.
PG?
-
@molnibalage said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@Mav-jp said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@Stevie said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@suhkoi69 - insider trading…
Bottom line is that it was more “correct” before it was “fixed”, and behaved more as I would expect.
no it was not
The previous modeling ( < 4.37) was a giant hack with absolutly no aquisition phase from the AIM120 seeker and with a PG of 100%
U3 improve the PG against non manoeuvering target, which was the major flaw of the previous implementation.
PG?
Probability of Guidance
-
@Mav-jp said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@molnibalage said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@Mav-jp said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@Stevie said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@suhkoi69 - insider trading…
Bottom line is that it was more “correct” before it was “fixed”, and behaved more as I would expect.
no it was not
The previous modeling ( < 4.37) was a giant hack with absolutly no aquisition phase from the AIM120 seeker and with a PG of 100%
U3 improve the PG against non manoeuvering target, which was the major flaw of the previous implementation.
PG?
Probability of Guidance
With that clarification, I have to ask: How does U3 “improve the PG against a non-maneuvering target” if the <4.37 was already 100%? When you say “improve” do you mean “make more realistic”?
-
@SoBad said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@Mav-jp said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@molnibalage said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@Mav-jp said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@Stevie said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
@suhkoi69 - insider trading…
Bottom line is that it was more “correct” before it was “fixed”, and behaved more as I would expect.
no it was not
The previous modeling ( < 4.37) was a giant hack with absolutly no aquisition phase from the AIM120 seeker and with a PG of 100%
U3 improve the PG against non manoeuvering target, which was the major flaw of the previous implementation.
PG?
Probability of Guidance
With that clarification, I have to ask: How does U3 “improve the PG against a non-maneuvering target” if the <4.37 was already 100%? When you say “improve” do you mean “make more realistic”?
no
<4.37 , the PG was always 100% whatever happened because there was no concept of missile seeker search phase, the Target Pointer was just passed from the FCR to the seeker , there was no actual search4.37 included a real search by the seeker at HPRF and MPRF, with errors build in the DL phase. One of the problem was that the PG had become too low against non manoeuvering target , because the Errors build on were too big but compensated by a search window that was too big (and therefore not selective enough)
4.37.U3 fixed the issue of PG beeing too low against non manoeuvering target, this is not back to 100% though , it has been done by reducing the amound of errors in DL phase (based on gaussian distribution) and a much reduce search window (based on the same gaussian distribution as well)
-
Also no longer possible to avoid a missile by heading straight towards it, nose down.
Zut alors.
-
I found it again: the straight line (with speed) is life.
It was fun and quite surprising for the opponent. -
@Rouge1512 mach >1.2 that involves a configuration without fuel tank…however, fuel = the life…
-
In general, I kill a lot of planes full of oil (which they have saved for nothing ) .
So speed : the life.
-
@Mav-jp - Thank you for that clarification, Mav. Forum posts/comments indicate that this is a worrisome issue for many people, and I’m sure they appreciate you taking the time to explain that.
-
@Mav-jp - the 4.35 model was/is still more like an actual missile behaves/works. What you describe of what you did just isn’t RL reality.
And I suspect that the APG-68 isn’t as resolute in elevation as I’d like to think…just looking at the antenna. In RL that factors too - it’s not just pu to the missile.
-
Just for clarification
4.37 was a total rewrite and 4.37.3 is another total re write (it’s not just a fix).
Falcon 4.0 => 4.36 :
No search phase from the missile seeker , FCR target pointer transmitted to the seeker automaticallyIssues :
PG was 100% whatever happened.
No possibility for the missile to search and guide to another target
Fake HPRF phase
Fantaisist Maddog phase
Fantaisist HOJ
Missile continuing guiding even with target out of FOV4.37 => 4.37.2
DL large errors build in , based on distances with a linear distribution
Huge window at search/ selection phase
Correct Maddog and HOJ phases
Missile trajectory limited by its FOVIssues
PG too low for non manoeuvering targets
PG should have been close to 100% with continuous FCR lock and it wasn’t the case (that was the biggest issue here )
Large deviations during flight => energy dissipation too big4.37.3 =>
DL small errors build in, based on angles with a Gaussian distribution
Small window at search / selection phase
Correct Maddog and HOJ phases
Missile trajectory limited by its FOVI encourage people to fly and test 4.37.3 and I’m sure you will all appreciate the new code
-
@Mav-jp thank you so much for clarifying and taking your time. helps a lot and also provides valuable insights. cheers
-
Excellent write-up as usual @Mav-jp, thank you. I have been doing testing and one thing I encounter is that the 120B will timeout at 80 seconds. I’m not sure if this is intended behavior (real world?) or not but it does mean that even though the shot may still be tracking and have enough energy to hit, it doesn’t because the missile self destructs. Any insights here?
The AGM-65G had a similar issue in U2 but I haven’t had a chance to test that yet.
Aside from that, my test under controlled conditions (TE with AN-24 for example) seem to validate your description of the changes mentioned above. Nice work!
-
@Zeus said in New implementation of ARH (Active Radar Homing logic for Fox3)?:
Excellent write-up as usual @Mav-jp, thank you. I have been doing testing and one thing I encounter is that the 120B will timeout at 80 seconds. I’m not sure if this is intended behavior (real world?) or not but it does mean that even though the shot may still be tracking and have enough energy to hit, it doesn’t because the missile self destructs. Any insights here?
The AGM-65G had a similar issue in U2 but I haven’t had a chance to test that yet.
Aside from that, my test under controlled conditions (TE with AN-24 for example) seem to validate your description of the changes mentioned above. Nice work!
The B always had a battery life of 80
For better battery life, fly the C