Path planning of mobile robots using genetic algorithms and modified artificial potential field

Dissertant

Abbas, Mustafa Wad

Thesis advisor

al-Shammari, Muhammad Jasim Muhammad

University

University of Technology

Faculty

-

Department

Department of Control and Systems Engineering

University Country

Iraq

Degree

Master

Degree Date

2013

English Abstract

Optimal global path planning for a mobile robot is considered the essential first step for realizing autonomous ground vehicles.

There are numerous methods and algorithms initially developed for a single mobile robot working in an environment containing static obstacles.

Among these methods and algorithms is the genetic algorithm.

This algorithm is an optimization and search technique which is inspirited from the principles of natural selection and natural genetics.

In this thesis, four algorithms are proposed to find the optimal path for mobile robot.

The first algorithm is “Genetic Algorithm Based on Cods of Nodes”.

In this algorithm, the environment is modeled by many nodes and links among them.

Proposed enhancements are added to the genetic operations to quick the convergence to the optimal path.

The second algorithm is “Enhanced Genetic Algorithm Based on Links among Distributed Nodes”.

This algorithm divided the space of environment into equally quarters by projecting a grid of specified distance on the environment space.

An obstacle is defined by one or more rectangular that surrounds it.

In this algorithm, many genetic and non genetic operations are proposed to increase the capability of this algorithm to find the optimal path.

The third algorithm is “Genetic Algorithm Based on Parameters of Artificial Potential Field”.

This algorithm uses the parameters of artificial potential field to deform this field to the form that results an optimal path.

The optimal parameters of this field are found by enhanced and modified proposed genetic algorithm.

The capability of this planning algorithm is increased by adding some proposed non genetic operations to facilitate the task of genetic algorithm.

The fourth algorithm is “The proposed algorithm for mobile robot obstacles avoidance”.

This algorithm is a local path planning algorithm, and it is proposed to adapt with any interrupt that could be occurred by unknown obstacle during the execution of the global path.

This algorithm is used to avoid obstacles in partially known environment based on the readings of five infrared sensors that are installed on robot with different angles relative to middle axis of the robot.

The philosophy of this algorithm is based on the modifications of artificial potential field technique.

Finally, in this thesis, the method of constructing the mobile robot from its assembly components is introduced in details.

This mobile robot is guided using remote computer through two types of interfaces techniques.

Moreover, our mobile robot was used to execute many global paths in the real world environment.

Main Subjects

Engineering & Technology Sciences (Multidisciplinary)

Topics

• Computing

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-419606