| Author |
Missile tracking |
Gejaheline
Fleet Admiral
Galactic Navy
Joined: March 19, 2005
Posts: 1127
From: UGTO MUNIN HQ, Mars
|  Posted: 2012-04-24 09:54
Now, as we all know, missiles currently leave something to be desired in the hitting-stuff department when really they should be at least capable of hitting stationary targets.
I'm reasonably sure someone's looking at it, but I thought I'd see if I can recruit some help for my own pet discussion and see if people can come up with some ideas.
The problem I've got, you see, is that I KNOW the principles behind guiding something to hit a target, but it's tricky to turn that into the formulae and programming required to actually guide a missile to hit the target. So if I hand out some of my notes, does anyone think they can build on them?
Okay, so. There are basically two ways to look at this.
The first one is by calculating the future impact point of the target and aiming the missile at that point. This is, I am fairly sure, how normal DS gunnery works and is what the current missile tracking system is based on.
This sounds pretty simple; take the speed of the target and the projectile, then take the distance, divide distance by projectile speed to get how long it'll take and then see how far the target will move in that time. Fire at that estimated point and you'll hit.
There are two complications, however. First off is that the time the projectile takes will be affected by how far away the aimpoint is. If they're moving away from you, the projectile will take longer to hit the aimpoint than it will to hit the target because it will have to move further. Which means you have to adjust the aimpoint. Which changes the flight time, which means you need to adjust the aimpoint, and so on. This requires lots of complicated sums.
The other complication is that it assumes the shot will travel at full speed directly towards the aimpoint. Missiles rarely if ever get fired directly at the target, meaning the predicted impact point will be wrong because the flight time will be different due to the missile taking extra time to follow a curved path.
Now, there's a second way to look at this. From the point of view of the target ship, it's really easy to see if something is going to hit you. If the object looks like it's travelling directly towards you, it'll hit you dead on. If it appears to be moving sideways, it's going to miss you. To put it in math-speak, if the angle betwen the nose of your ship and the target is not changing AND the distance is decreasing, that object WILL hit you. Angular velocity will be zero.
As it happens, this holds completely true for the projectile, as well. This is how semi-active-homing missiles work: They maintain a steady angle compared to a reflected signal from the target.
This is quite obvious if the target isn't moving. If you point your nose at a stationary ship and fly forwards, your camera view doesn't move and you run into the ship. If you're not pointing straight at them and are drifting to one side or the other, your camera will slowly pan to keep the target centred in your screen and you'll miss them.
It's the same principle with a moving target, but you have to compensate for their movement.
So, the challenge here (and I'll be working on it myself) is to come up with a method to use one or both of these facts to guide a missile to the target. I'm not asking anyone to do any programming or anything, mind, just a description or a formula or whatever will help.
Disclaimer: I am not a dev, I am not asking for help on behalf of Palestar, this post if for entertainment purposes first and foremost.
_________________
[Darkspace Moderator] [Galactic Navy Fleet Officer]
|
Pantheon
Marshal
Palestar
Joined: May 29, 2001
Posts: 1789
| Posted: 2012-04-24 10:10
This will be a dev decision only. You can theorise your calculations to the nth degree, but we're only interested in server performance, which is why tracking is how it is now.
[ This Message was edited by: Pantheon on 2012-04-24 10:11 ]
_________________
|
CM7
Midshipman
Faster than Light
Joined: October 15, 2009
Posts: 1812
| Posted: 2012-04-24 10:15
you need a combo of the two.
First the fire control will give an xy cord for interception of the target.
Next you gave a cone of detection originating from the nose of the missile out to the max range of said missile. Diamiter of the base of the cone can be adjusted per missile.
Once you fire the projectile, it turns to its programmed intercept cords. When yhe target is picked up in the detection cone, the missile turns to hold the target in the cone, overiding the original flightplan. As the projectile comes closer to the target, the diamiter of the walls of the cone get smaller like a funnel.
1. Intercept
2. Detect
3. Track
4. Adjust
_________________
Defiance and Opposition, a tribute to teamwork. I will remember always
339,144
|
Forger of Destiny
Chief Marshal
We Kick Arse
Joined: October 10, 2009
Posts: 826
| Posted: 2012-04-24 11:10
i like using ways which are already there, just not utilised.
harpex missiles shot at weapons platform from 700 gu range. missile misses.
i notice, that as a missile aims its flight direction at the target, it's rate of aiming gets slower. i can explain this
missile - ok, weapon platform 30 degrees above me, i should turn up
missile turns at max turn rate of 15 degrees/second, platform 500 gu away.
missile - weapon platform 15 degrees above me, ill definitely hit it. nothing will go wrong if i turn slowly.
missile turns at 7.5 degrees/second, platform 400 gu away.
missile - hmm, platform is supposed to be only 7.5 degrees off my vector, why is it at 10 degrees?
missile turns 5 degrees/second, platform 300 gu away.
missile - platform is still 8 degrees off my vector
missile keeps turning slowly, until it totally misses target and either gets shot or tries to turn around. 24 seconds and ATLEAST 360 degrees later, missile misses again and gets banned by server for misappropriate CPU utilisation.
this "dynamic turn rate" may be nice and all, but i feel its better for a missile to turn 30 deg/s for 0.2 seconds and be aimed accurately, than to turn 6 deg/s for an entire second and miss anything smaller than a dread.
i feel missiles miss moving targets because they compensate their flight path and take into account the speed of target.
or, they compensate too much.
for moving targets, missiles should aim directly at the center of the ship - no compensating. this, combined with a no-compromise turning system will greatly increase missile accuracy.
1) better turning rate
2) no course correction for target's motion vector
without accounting for the target's motion, missiles would be able to be shot at targets running out of missile range. missile wasted?
i'd say missile "shooting" range should be reduced to 75-80% of current range, and missile speed be increased 10% (no change to missile lifetime). this way, missiles can chase targets with their great 100+ gu/s speeds, and nail even a speeding scout.
3) reduce missile targeting range
4) increase missile speed
byebye
[ This Message was edited by: Toby D Syded on 2012-04-24 11:30 ]
_________________
Forging legends and lives outside till naught remains inside.
|
CM7
Midshipman
Faster than Light
Joined: October 15, 2009
Posts: 1812
| Posted: 2012-04-24 15:27
Quote:
|
On 2012-04-24 10:10, Pantheon wrote:
This will be a dev decision only. You can theorise your calculations to the nth degree, but we're only interested in server performance, which is why tracking is how it is now.
[ This Message was edited by: Defiance*CO* on 2012-04-24 15:37 ]
|
|
your...
She said this is not for the dev team. Its just a fun activity.
_________________
Defiance and Opposition, a tribute to teamwork. I will remember always
339,144
|
Talien
Marshal
Templar Knights
Joined: May 11, 2010
Posts: 2044
From: Michigan
| Posted: 2012-04-24 16:14
Quote:
|
On 2012-04-24 11:10, Toby D Syded wrote:
this "dynamic turn rate" may be nice and all, but i feel its better for a missile to turn 30 deg/s for 0.2 seconds and be aimed accurately, than to turn 6 deg/s for an entire second and miss anything smaller than a dread.
|
|
That's what missiles are designed to do. They're primarily for use against DN/Stations, if you're trying to shoot anything smaller you shouldn't be upset when you realize they aren't hitting.
As far as the arc making them miss things under 700 GU distance, do you want to avoid FF with an arc or do you want to have a minimum distance you know you're probably not going to hit anything under, which includes allies?
_________________
Adapt or die.
|
Lithium
Chief Marshal
Joined: June 29, 2003
Posts: 109
| Posted: 2012-04-24 17:06
Missiles miss if target is too close.
That's why missiles retargetted don't hit new target 
Missile path should be wider if target is too close.
_________________
|
Fluttershy
Fleet Admiral
Joined: September 24, 2011
Posts: 778
From: Fluttershy
| Posted: 2012-04-24 17:56
Why don't you just get rid of missiles altogether, and replace them with bombardment cannons that have reverse falloff? (weak up close, strong at range)
_________________
|
DiepLuc
Chief Marshal
Joined: March 23, 2010
Posts: 1187
| Posted: 2012-04-25 03:09
Quote:
| On 2012-04-24 09:54, Gejaheline wrote:
Now, there's a second way to look at this. From the point of view of the target ship, it's really easy to see if something is going to hit you. If the object looks like it's travelling directly towards you, it'll hit you dead on. If it appears to be moving sideways, it's going to miss you. To put it in math-speak, if the angle betwen the nose of your ship and the target is not changing AND the distance is decreasing, that object WILL hit you. Angular velocity will be zero. |
|
In beta, I fired missles from Ganglia to a Claw at 400gu. All missed! The claw didn't move. But they hit the cruiser at the same range.
This is why Lith said:
Quote:
| On 2012-04-24 17:06, Lithium wrote:
Missiles miss if target is too close. |
|
_________________
|
Iwancoppa
Fleet Admiral
Joined: November 15, 2008
Posts: 709
| Posted: 2012-04-25 03:30
Allthough unrealistic, possibly missiles could work by mearly homing in onto the position of the enemy ship and if they miss, overshoot the target to their max turning diameter + 100gu and then turn around and try again.
_________________
|
Fluttershy
Fleet Admiral
Joined: September 24, 2011
Posts: 778
From: Fluttershy
| Posted: 2012-04-25 04:28
What's unrealistic about that?
_________________
|
Blackjack [DBL]
Grand Admiral
Faster than Light
Joined: February 25, 2011
Posts: 344
From: The land of venomous reptiles.
| Posted: 2012-04-25 04:58
Quote:
|
On 2012-04-25 03:30, iwancoppa wrote:
Allthough unrealistic, possibly missiles could work by mearly homing in onto the position of the enemy ship and if they miss, overshoot the target to their max turning diameter + 100gu and then turn around and try again.
|
|
Some modern Fire-and-forget missiles are designed to be able to do that. It actually exists.
_________________
Names I used: Da Bes Loser, Perseverance, Loyalty.
|
Gejaheline
Fleet Admiral
Galactic Navy
Joined: March 19, 2005
Posts: 1127
From: UGTO MUNIN HQ, Mars
| Posted: 2012-04-25 05:36
It's rather difficult to say "that's unrealistic" because we're talking about inherently unrealistic space-missiles, but a pure-pursuit curve (where you point your nose at the target) is in fact used in real air combat for achieving missile locks and making it harder for the target to take evasive action.
For missiles it tends to get used in active homing missiles, where the missile itself has a radar mounted in the nose to seek the target; basically it just points itself directly at the biggest, closest blob it can see. This is perfectly acceptable for medium-ranged missiles because the missile's sheer speed outweighs any real need to lead the target.
Other guidance systems that include datalinks or launcher-based radar can perform cleverer tricks like guiding the missile to where the target is going to be, either by having the missile take a bearing off a reflected radar beam or sending the missile direct steering commands.
Meanwhile, I've been wrestling with trigonometry, which is slowly driving me insane because it does things like give the same result for two different sets of coordinates and send my simulated missiles careening off in the opposite direction to the target because the sine function can't tell the difference between "up and to the left" and "up and to the right", auuugh. Any trig experts in the house?
_________________
[Darkspace Moderator] [Galactic Navy Fleet Officer]
|
DiepLuc
Chief Marshal
Joined: March 23, 2010
Posts: 1187
| Posted: 2012-04-25 06:16
A simple solution for missle is to increase its blast range and make it auto explode when the target is in blast range.
Normal beam range is 150gu... so light, medium heavy missle radius can be 50gu, 90gu, 120gu respectively...
I dislike missle seft destruct when target disappear. It applies to all flying pods generally. It'd be nice if the flying pod updates the coordinates of the target every 3 seconds and just go to the last location when the target out of sensor range.
_________________
|
Enterprise
Chief Marshal
Joined: May 19, 2002
Posts: 2576
From: Hawthorne, Nevada
| Posted: 2012-04-25 08:00
I would personally prefer that things like Missiles become massive area of effect weapons.
Essentially, they would have poor tracking, but have incredible AoE, yet, up to -90% damage at the edge, to full damage within close range.
This way, we don't have to focus so much on tracking, as on balancing the damage. This way also, missiles become effective cluster damage weapons without the unneccessary processing power that goes into tracking.
Given my personal preferences, I also wish torpedos has a much slower recharge with much less ammo and take much more energy, and become powerful anti-captial ship weapons. I think the reasoning behind high Dreadnaught usage has outlived its appeal.
-Ent
[ This Message was edited by: Enterprise on 2012-04-25 08:03 ]
_________________
|
English
Deutsch
Nederlands
)
