UDKBase.UDKBot

Extends

AIController

Modifiers

native

AI Controller that simulates the behavior of a human player Copyright 1998-2011 Epic Games, Inc. All Rights Reserved.

Core.Object
|   
+-- Engine.Actor
   |   
   +-- Engine.Controller
      |   
      +-- Engine.AIController
         |   
         +-- UDKBase.UDKBot

Direct Known Subclasses:

UTBot

Functions Summary
	float	AdjustAimError (float TargetDist, bool bInstantProj)
		BuildSquadRoute ()
	bool	CanMakePathTo (Actor A)
		DelayedLeaveVehicle ()
		DelayedWarning ()
		ExecuteWhatToDoNext ()
	Actor	FaceActor (float StrafingModifier)
	actor	FindBestInventoryPath (out float MinWeight)
	Actor	FindBestSuperPickup (float MaxDist)
	actor	FindPathToSquadRoute (optional bool bWeightDetours)
		LatentWhatToDoNext ()
		MayDodgeToMoveTarget ()
		MissedDodge ()
		MonitoredPawnAlert ()
		ReceiveRunOverWarning (UDKVehicle V, float projSpeed, vector VehicleDir)
	bool	SpecialJumpCost (float RequiredJumpZ, out float Cost)
	float	SuperDesireability (PickupFactory P)
		TimeDJReset ()
		WaitToSeeEnemy ()
		WhatToDoNext ()

AcquisitionYawRate Source code

The bot uses AcquisitionYawRate instead of RotationRate to track enemy before acquired.

AimUpdateFrequency Source code

bAllowRouteReuse Source code

if true, when pathfinding to the same RouteGoal as the last time, use old RouteCache if it's still valid and all paths on it usable

BaseTrackingReactionTime Source code

How far back in time is bots model of enemy position based on

bEnemyAcquired Source code

bEnemyAcquired is true if the bot has been able to face the enemy directly. The bot has improved tracking after acquiring an enemy (uses AcquisitionYawRate instead of RotationRate to track enemy before acquired).

bEnemyInfoValid Source code

bEnemyIsVisible Source code

bExecutingWhatToDoNext Source code

set when in ExecuteWhatToDoNext() so we can detect bugs where it calls WhatToDoNext() again and causes decision-making to happen every tick

bForceRefreshRoute Source code

used with route reuse to force the next route finding attempt to do the full path search

bInDodgeMove Source code

bInDodgeMove is true if a bot is currently executing a dodge, and has not yet landed from it

bJumpOverWall Source code

bLeadTarget Source code

Result of last enemy LineOfSightTo() check

BlockedAimTarget Source code

Iterative aim correction in progress if set

bNeedDelayedLeaveVehicle Source code

triggers the bot to call DelayedLeaveVehicle() during its next tick - used in the 'non-blocking' case of LeaveVehicle()

bPlannedJump Source code

bScriptSpecialJumpCost Source code

script flags that cause various events to be called to override C++ functionality

bTargetAlternateLoc Source code

if set pass bRequestAlternateLoc = TRUE to GetTargetLocation() when determining FocalPoint from Focus

bUsePreviousSquadRoute Source code

if true, this bot uses the SquadAI's PreviousObjectiveRouteCache instead (used when the route changes while bot is following it)

bUsingSquadRoute Source code

whether bot is currently using the squad alternate route - if false, FindPathToSquadRoute() just calls FindPathToward(Squad.RouteObjective)

CurrentAimError Source code

aim error value currently being used

CurrentlyTrackedEnemy Source code

Normally the current enemy. Reset SavedPositions if this changes.

DecisionComponent Source code

component that handles delayed calls ExecuteWhatToDoNext() to when triggered

DodgeLandZ Source code

expected min landing height of dodge. Hint used by bots to track whether they are about to miss their planned landing spot (if bInDodgeMove is true).

DodgeToGoalPct Source code

Likelihood that MayDodgeToMoveTarget() event is called when starting MoveToGoal().

EnemyVisibilityTime Source code

ErrorUpdateFrequency Source code

how often aim error is updated when bot has a visible enemy

FearSpots[2] Source code

avoid these spots when moving - used for very short term stationary hazards like bio goo or sticky grenades.

HearingThreshold Source code

distance at which bot will notice noises.

ImpactVelocity Source code

velocity added while falling (bot tries to correct for it)

LastAimUpdateTime Source code

Last time tracking aim was updated

LastBlockingVehicle Source code

Last vehicle which blocked the path of this bot

LastErrorUpdateTime Source code

last time aim error was updated

LastIterativeCheck Source code

pct lead for last targeting check

LastSeeingPos Source code

LastSeenPos Source code

LastSeenTime Source code

MonitoredPawn Source code

MonitorMaxDistSq Source code

MonitorStartLoc Source code

for monitoring the position of a pawn. Will receive a MonitoredPawnAlert() event if the MonitoredPawn is killed or no longer possessed,

MultiJumpZ Source code

jump Z velocity bot can gain using multijumps (not counting first jump)

RespawnPredictionTime Source code

How long ahead to predict inventory respawns

SavedPositions Source code

Position and velocity of enemy at previous ticks.

ScriptedFireMode Source code

if not 255, bot always uses this fire mode - for scripting bot actions

Squad Source code

Squad this bot is in. Squads share the same goal, and handle AI planning.

SquadRouteGoal Source code

goal along squad's route, used when moving along alternate path via FindPathToSquadRoute()

TemporaryFocus Source code

temporarily look at this actor (for e.g. looking at shock ball for combos) - only used when looking at enemy

TrackedVelocity Source code

Base value, modified by skill to set TrackingReactionTime

TrackingReactionTime Source code

Enemy tracking - bots use this for targeting their current enemy. Bots actually aim at a position based that's predicted based on the target's position TrackingReactionTime ago, extrapolated using the target's velocity at that time. This means, for example, that if a target suddenly changes directions, bots will miss because they'll fire at where they thought he was going

VisibleEnemy Source code

When last enemy LineOfSightTo() check was done

WarningDelay Source code

delay before act on warning about being shot at by InstantWarnTarget() or ReceiveProjectileWarning()

WarningProjectile Source code

Projectile bot has been warned about and will react to (should set in ReceiveProjectileWarning())

EnemyPosition Source code

AdjustAimError Source code

BuildSquadRoute Source code

Called by squadleader. Fills the squad's ObjectiveRouteCache. Builds upon previous attempts by adding cost to the routes in the SquadAI's SquadRoutes array, up to a maximum of MaxSquadRoutes iterations.

CanMakePathTo Source code

assumes valid CurrentPath, tries to see if CurrentPath can be combined with path to A

DelayedLeaveVehicle Source code

called just before the AI's next tick if bNeedDelayedLeaveVehicle is true

DelayedWarning Source code

ExecuteWhatToDoNext Source code

entry point for AI decision making this gets executed during the physics tick so actions that could change the physics state (e.g. firing weapons) are not allowed

FaceActor Source code

Used to determine which actor should be the focus of this bot (that he looks at while moving)

FindBestInventoryPath Source code

Searches the navigation network for a path that leads to nearby inventory pickups.

FindBestSuperPickup Source code

Sets all available "super pickups" as possible destinations, and determines best one based on the value of and distance to the pickup (distance must be less than MaxDist). A super pickup is a PickupFactory with bIsSuperItem==true To be valid, it must be ready to pick up, or about to respawn (within this bot's RespawnPredictionTime) not blocked by a vehicle, and desireable to this bot (based on the SuperDesireability() event).

FindPathToSquadRoute Source code

Returns shortest path to squad route (UTSquadAI.ObjectiveRouteCache), or next node along route if already on squad route

LatentWhatToDoNext Source code

encapsulates calling WhatToDoNext() and waiting for the tick-delayed decision making process to occur using UDKAIDecisionComponent

MayDodgeToMoveTarget Source code

MissedDodge Source code

If DodgeLandZ (expected min landing height of dodge) is missed when bInDodgeMove is true, the MissedDodge() event is triggered.

MonitoredPawnAlert Source code

Will receive a MonitoredPawnAlert() event if the MonitoredPawn is killed or no longer possessed, or moves too far away from either MonitorStartLoc or the bot's position

ReceiveRunOverWarning Source code

Warning from vehicle that bot is about to be run over.

SpecialJumpCost Source code

Called when bScriptSpecialJumpCost is true and the bot is considering a path to DestinationActor that requires a jump with JumpZ greater than the bot's normal capability, but less than MaxSpecialJumpZ calculates any additional distance that should be applied to this path to take into account preparation, etc
@return true to override the cost with the value in the Cost out param, false to use the default natively-calculated cost

SuperDesireability Source code

TimeDJReset Source code

If bot gets stuck trying to jump, then bCanDoubleJump is set false for the pawn Set a timer to reset the ability to double jump once the bot is clear.

WaitToSeeEnemy Source code

return when looking directly at visible enemy

WhatToDoNext Source code

triggers ExecuteWhatToDoNext() to occur during the next tick this is also where logic that is unsafe to do during the physics tick should be added @note: in state code, you probably want LatentWhatToDoNext() so the state is paused while waiting for ExecuteWhatToDoNext() to be called

defaultproperties
{
	RotationRate=(Pitch=30000,Yaw=30000,Roll=2048)
	RemoteRole=ROLE_None

	Begin Object Class=UDKAIDecisionComponent Name=TheDecider
	End Object
	DecisionComponent=TheDecider
	Components.Add(TheDecider)

	AimUpdateFrequency=0.2
	LastIterativeCheck=1.0
	ErrorUpdateFrequency=0.45
	bLeadTarget=True

	AcquisitionYawRate=20000
	TrackingReactionTime=+0.25
	BaseTrackingReactionTime=+0.25
	bUsingSquadRoute=true
	bAllowRouteReuse=true
	HearingThreshold=2800.0
}