Abstract:
This project is an attempt to find a solution through a two-dimensional labyrinth of obstacles which can be described as polygons. The optimal solution can be found through extensive computation, however through utilizing a genetic algorithm it is possible to find an acceptable or “close enough” solution. A genetic algorithm uses the concept of natural selection to narrow down a list of possible solutions, and over generations evolve a closer and closer solution, combining or mutating good solutions for future generations. With a smaller number of solutions that are checked quickly and narrowed down, it is possible to drastically cut down on computation time that it takes to find a working solution.
This project is an attempt to find a solution through a two-dimensional labyrinth of obstacles which can be described as polygons. The optimal solution can be found through extensive computation, however through utilizing a genetic algorithm it is possible to find an acceptable or “close enough” solution. A genetic algorithm uses the concept of natural selection to narrow down a list of possible solutions, and over generations evolve a closer and closer solution, combining or mutating good solutions for future generations. With a smaller number of solutions that are checked quickly and narrowed down, it is possible to drastically cut down on computation time that it takes to find a working solution.
CSCI 460: Capstone Experience Project Definition
2019 Pankratz/McVey
Path Planning with Obstacles
Project Developer: Wyatt Verheyen
Genetic algorithms can be used to generate multiple solutions to a given problem. Finding the optimal path in a space filled with obstacles can be found exactly but not without extensive computation. A genetic algorithm can be used to generate samples until an acceptable, but perhaps not optimal, path is found.
Project Description: Develop an application that discovers an acceptable path through a factory floor with fixed obstacles.
General Requirements:
2019 Pankratz/McVey
Path Planning with Obstacles
Project Developer: Wyatt Verheyen
Genetic algorithms can be used to generate multiple solutions to a given problem. Finding the optimal path in a space filled with obstacles can be found exactly but not without extensive computation. A genetic algorithm can be used to generate samples until an acceptable, but perhaps not optimal, path is found.
Project Description: Develop an application that discovers an acceptable path through a factory floor with fixed obstacles.
General Requirements:
- Graphically create and display a factory floor with polygonal obstacles, a source, and a destination.
- Use a genetic algorithm to find an acceptable path from a given source to a specified destination on the factory floor.
- Graphically display the state of good fit genomes with appropriate statistics that show progress toward the goal.
- The system should allow for modification of the genome scoring function.
- The system should allow for modification of parameters, operators, and probabilities.