J. Savage and E. Marquez (Mexico)
Mobile robots, Genetic algorithms
In this paper we describe a learning module of a mobile robot’s architecture, that consists of several layers that control the operation of real and virtual robots. Each layer has many modules, and one of the most important module in the architecture is the one concerning with learning. Learning helps to improve the robots navigation performance by finding navigation landmarks which guide the mobile robot from an origin to a destination. There are several algorithms that find the best path or a path given an origin, a destination and a symbolic representation of the environment, one way to find a solution is by using genetic algorithms (GA). In our approach each chromosome is represented by a set of points, that represents landmarks, that the robot needs to follow from an origin to a destination. Each individual belongs to a population, and it is evaluated according to a fitness function that evaluates how good are the landmark points that take the robot from an origin to a destination. Our objective was to prove that GA is a good option as a method for finding landmarks for mobile robots’ navigation
Important Links:
Go Back
