Trajectory Question

[jellytot]

Hi all,

I have a question. This is the setup:

There is a Target on screen. It is moving across the screen in a straight line, at some angle (dirT) at some fixed speed (speedT).
There is a Player object on screen. It is not moving. On a mouse click, it can fire a Missile.
A Missle moves in a straight line, at some angle (dirM) at some fixed speed (speedM).

What I want to do, is when I click the mouse the game generates a Missile at the Player object, it knows the DIRECTION (dirM) to set the Missile at, so the Missile (if possible) will intercept/collide with the Target. Keep in mind both the Missile and Target are moving in a straight line, and possibly at different speeds.

dirM and dirT are 360 deg.

Does anyone know how to do this?

JT

[Gerblegod7]

Hmmmmmmm, this would be solved with a linear system. I'm fairly new to the concept and wouldn't know how to implement it in MMF2.

[SEELE]

Add
sin(angle)*speed to alterable value A of missile
Add
cos(angle)*speed to alterable value B of missile

set missile x to alterable value a of missile
set missile y to alterable value b of missile


The alterable value step is unfortunately necessary because positions are cannot hold decimal values and integers would not move smoothly.

[jellytot]

Thanks,

A few questions,

How do I return an object's angle? I tried using the Advanced Direction object but couldn't figure it out.

Come to think of it, how do I SET an object's angle? Isn't MMF limited to 32 directions?

JT

[SEELE]

No.

In the advance direction object go retrieve angle between two points and put in x1,y1,x2,y2.

Or set angle to sqr((x2-x1)pow 2)+((y2-y1)pow 2))

[jellytot]

I couldn't figure out which 2 objects to use to get angle.
I tried doing all variations of Target/Missile/Player.

Also, it seems that the formula calculates the dynamic position assuming the Target does not change direction. Is there a formula that calculates the direction (or position) for Missile to intercept Target at the time that Missile is fired? That way, if Target changes direction Missle would have intercepted target had it not changed direction.

JT

[Dines]

sqr((x2-x1)pow 2)+((y2-y1)pow 2))

That won't get the angle, that's the distance. To get the angle, you'll need an extension. The Clickteam Movement Controller can do it.

Jellytot: So you want a system to calculate what trajectory a missile would have to take if it wanted to intercept an object, assuming that object will not deviate in its speed or direction?

I've got a working example, but it's not quite what you'd hope for.

A) It's actually quite tough to pull off (at least my way is, maybe someone can do a simpler method),
B) It may not be able to hit the target unless you can vary the speed of the missile.

The problem is that if the missile goes slower than the target, it's possible for the target to outrun the missile. At best, my system could in this circumstances tell you how far you'd miss it by, which could be used to increase the speed so that you could hit it.

The system can also tell you the accuracy with which it will fire.

How it works is thus:

There's a target at (tX,tY), moving at tSpeed towards tAngle.

The player fires the missile. The missile is created, and a fastloop starts. This fastloop will be used as a kind of virtual reality, or imagination, for the missile to predict where it and the target will be.

The missile uses the following alterable values:

Velocity (its speed)
Movement Angle (what direction it'll fire in)
XPOS (its own X position)
YPOS (its own Y position)
Predicted XPOS of target (where it imagines the target will be)
Predicted YPOS of target (where it imagines the target will be)
Predicted Self Distance (how far the missile itself will have travelled)
Current Lowest Distance (defaults to 9999999)


At start, the missile loads the target coordinates into its predictive XPOS and YPOS slots. In every loop that follows, it adds the target's velocity to those coordinates (using trigonometry to adjust for angle).


At the same time, it keeps adding its own velocity to the Self Distance variable.

So, basically, each loop will represent one frame in the future. So after 25 loops, it's looking 25 frames into the future.

These vars enable the missile to 'imagine' where the target will be in X frames time, and compare it to how far it itself will have travelled.

On each loop, it compares the distance it will have travelled with the distance between missile starting point and the predicted location of the target.

It selects the one with the lowest distance. Then, it calculates the angle it needs to take (the angle between the missile's origin point and the predicted future location of the target).

The loop finishes, the missile launches, and a few moments later, they colide.

It works, but like I said, if the missile is slower than the target, then there are no guarantees that it'll hit its mark.

[Nifflas]

You don't need to use an extension to compare the distance between two points though. You can either use this or this method. Using an extension is a bit easier and more CPU friendly, but adding lots of extensions can potentially make your application unstable.

[Dines]

THIS should show a working example.

Remember that if you update the values of the objects so that they go very fast, or that the missile is slower than its target, there will be situations where the two don't meet.

In the first case, it's because the two would go so fast they 'skip over' each other. (edit: obviously, you'd need to do an in-between movement to avoid this problem).

In the second case, there will be some angles where the target simply pulls away from the bullet.

[jellytot]

Wow! Thanks so much!

JT