IGV_Simulator

Simulates the Motion of the IGV

Pathfinding..

--

Project Details:

• Simualator contains 2D map of environment..
• Environment contains: Bot, Obstacles, GPS_Waypoints
• Bot needs to be able to move through the simulator; Pathfind from Start to Target
• Bot contains: It's own information about the environment..

Problems:

• How can we describe objects in the map?
• How do we find the list of nearby objects we are interested in?
• What methodology is needed to find a path?
• How can we be sure that our path is drivable? (A Path is one thing, the physical limitations of the IGV are another..)
• How do we define drivable? (ie: max turn radius at given speeds, not enough space, ..)

--

• Bot has Search Radius. When an obstacle enters the Bot radius.. the Bot must store the position of said obstacle
• in it's internal map. (The Obstacle has it's own properties/functions that describe it's surface)
• When we need to generate the Bot's Path, the Bot needs to consider Obstacles in it's map
• When we need to identify obstacles.. we need to make a prediction given the information about the obstacle..

Solutions:

1. Points, whose shape covers a discrete surface area of the map.. stored in the Grid Map System
2. Using IGV(x,y) .. index into Grid Map.. +/- searchRadius
3. A* -> store/update path in graph structure (of paths) .. based on IGV's Discovered Environment
4. JSmoke libraries :)
5. (Define Limitations.. apply them to generating a close to optimal path.. maybe use some sampling techniques based on previous turning experience.. aka: we would have to train the bot by manually driving it... and generate Weights for turns and motion.. etc.. then we come back here and use our collected data...)

Ideas:

Simulator:

Map contains:

• • IGV
• • Obstacles

    Logic contains:

• • Collisions
• • Drawing stuff onto screen

IGV

Map contains:

• ( generated as exploration continues.. IGV update's its location and it's environment in it's own map of the world )
• • IGV
• • Obstacles
• • Waypoints

Graph contains:

• • Pathing data.. (relative/(and based solely on)/to the Map it has generated..)

Logic contains:

• • Driving, (updating it's location on the Simulator map..)
• • Update It's own map by discovering Surroundings.
• • Determine Pathing_Data given IGVs Map..
• • Determine the likeliness this path is Correct, given Limitations and Error of Data..
• • Make corrections and chose better path -> so the that IGV doesnt crash and burn..
• (better is not more optimal.. it is just Drivable..)

Please make changes to Ideas that need be updated. (for: accuracy & feasability)