Modular System to Train and Test an Evolutionary Mobile Robot for Obstacle Avoidance

Main Article Content

Tefool H. O. Al-khafaji
Ali Hasan Shaheed

Abstract

The field of evolutionary robotics is an interesting research area that deals with building neuro-controller systems by evolving individuals depending on the Darwinian principles of natural selection and survival of the fittest. It stems its importance from the need of building an intelligent robots that can learn to behave autonomously in unexpected situations in unknown and unpredictable environments.


Materials and Methods:


This paper produces Breeder Genetic Algorithm (BGA) as the tool in an evolutionary mobile robot system. BGAs share aspects with traditional GAs and evolution strategies. The controller is tested in the task of obstacles avoidance. Also the system is tested when two tasks are wanted to be modulated (Homing and obstacle avoidance).


Results:


It is shown that evolution enhances the performance of the system in terms of the average population fitness and the best individual fitness. It is also demonstrated that it is feasible to use several different modules that can cooperate to perform a given task. 


Conclusion:


Different runs in the evolutionary systems can produce different solutions. Also, simulations can help making several important decisions concerning the initial parameters governing the evolutionary controller systems. The fitness function must properly be designed to yield the desired behavior. All the experiments exhibit an enhancement in the level of the average population fitness and best individual fitness across generations. Furthermore, the performance of evolved individuals can be evaluated subjectively by observing the trajectories and behavior made by the best controller developed by the BGA.

Article Details

How to Cite
[1]
“Modular System to Train and Test an Evolutionary Mobile Robot for Obstacle Avoidance”, JUBPAS, vol. 32, no. 1, pp. 25–44, Mar. 2024, doi: 10.29196/g6w3s846.
Section
Articles

How to Cite

[1]
“Modular System to Train and Test an Evolutionary Mobile Robot for Obstacle Avoidance”, JUBPAS, vol. 32, no. 1, pp. 25–44, Mar. 2024, doi: 10.29196/g6w3s846.

Similar Articles

You may also start an advanced similarity search for this article.