Presentation is loading. Please wait.

Presentation is loading. Please wait.

12/19/2003Cat Hunt: Expanded Dog Chasing Cat Problem 1 Cat Hunt Expanded Dog Chasing Cat Problem By Eric Olson.

Similar presentations


Presentation on theme: "12/19/2003Cat Hunt: Expanded Dog Chasing Cat Problem 1 Cat Hunt Expanded Dog Chasing Cat Problem By Eric Olson."— Presentation transcript:

1 12/19/2003Cat Hunt: Expanded Dog Chasing Cat Problem 1 Cat Hunt Expanded Dog Chasing Cat Problem By Eric Olson

2 Cat Hunt: Expanded Dog Chasing Cat ProblemEric Olson 2 12/19/2003 Problem Expand and improve the dog-cat problem  Increase the number of dogs and giving them improved A.I. in chasing the cat  Give the cat A.I. to aid it evading the dogs Create clearer and easier to use graphical representation of chase  Remove animal trails  Static axes

3 Cat Hunt: Expanded Dog Chasing Cat ProblemEric Olson 3 12/19/2003 Motivation It was relatively boring to watch the cat run straight and have the dog circling it. Cats are smart, let it try to escape. Dogs are smarter than the original simulation. Dogs really don’t like cats, get more dogs!

4 Cat Hunt: Expanded Dog Chasing Cat ProblemEric Olson 4 12/19/2003 Solution: Setup Cat Properties  Only one cat  Twice as fast as dogs  Initial location is always in center of area (0,0) Dog Properties  Can be unlimited dogs  Initial location(s) randomly generated or loaded from file Simulation Ends When:  A dog is within.5 units of cat  Cat escapes off of area 600x600 square  Cat eludes dogs for 500 iterations

5 Cat Hunt: Expanded Dog Chasing Cat ProblemEric Olson 5 12/19/2003 Solution: Cat A.I. Cat bases route off of closest dog  Determine closest dog  Run away from closest dog at angle it is pursuing  Define route after every iteration of system

6 Cat Hunt: Expanded Dog Chasing Cat ProblemEric Olson 6 12/19/2003 Solution: Dog A.I. Dogs base route off of current and possible future locations of the cat Steps to determine route of each dog  Determine distance from cat  If cat is more than 5 steps (iterations away), run towards where the cat should be then if no course change is made.  Otherwise, run directly at where the cat should be the next iteration.

7 Cat Hunt: Expanded Dog Chasing Cat ProblemEric Olson 7 12/19/2003 Graphical Modes Routes with TrailsRoutes without Trails

8 Cat Hunt: Expanded Dog Chasing Cat ProblemEric Olson 8 12/19/2003 Test Results: Max-Min vs. Kosko’s Max-Product Rule # DogsTrials% Caught% Escaped 11000 2 199 31001486 41001387 51002476 81004654 101005545 151006040 201007228 251007525 301007723 401008515 501008911 100 982 # DogsTrials% Caught% Escaped 11000 2 0 3 496 41001585 51002476 81003763 101005347 151006238 201007228 251006832 301007030 401008218 501008416 100 982 Max-Min RuleKosko’s Max-Product Rule

9 Cat Hunt: Expanded Dog Chasing Cat ProblemEric Olson 9 12/19/2003 Test Results: Max-Min vs. Kosko’s Max-Product Rule

10 Cat Hunt: Expanded Dog Chasing Cat ProblemEric Olson 10 12/19/2003 Conclusion Expansion makes for a very interesting simulation Random dog locations increases intrigue as each new test is different Extremely fun project to complete Interesting application of Fuzzy Control Logic


Download ppt "12/19/2003Cat Hunt: Expanded Dog Chasing Cat Problem 1 Cat Hunt Expanded Dog Chasing Cat Problem By Eric Olson."

Similar presentations


Ads by Google