Reference Number: INTAS-97-2028
INTAS logo Intelligent Neural System for Autonomous Control of a Mobile Robot
Summary of results
Home | Participans | Objectives | Expected results | Publications
Research activities | Scientific results | Summary of results
Summary of results
As a result of this two-years project the Intelligent neural system for the control of a mobile robot was developed. Such Intelligent System is based on Artificial Neural Networks and capable of performing the different aspects of autonomous driving, such as collision-free motions, avoiding obstacles, mapping and planning of path.
As sensor inputs, providing the source information for such a control system is used the following data:
  • Video image of the path in front of the vehicle,
  • Range sensors to detect obstacles
  • Dead reckoning of location
On the basis of this data, the neural system solves the following tasks:
  • Data fusion
  • Reactive control of the mobile robot while moving in the unknown environment
  • Formation of the global route map during driving in the unknown area
  • Planning of an optimum route for the movement

Neural system has ability for robust control in case of inexact data from sensors, supervised and unsupervised training, real-time action. Self-training is realized both on the reactive level and on the level of path planning.

The result of the project is the software for control of a mobile robot, which experimentally was tested on the mobile robots. The experiments were performed using mobile robots "Walter" and "Pioneer" at FH Ravensburg-Weingarten. They showed capability for real-time reactions and good confirmation to the theoretical results. The robot was able to pass through narrow openings (e.g. doorways) or narrow corridors without collisions using inexact data from sensors.

The developed Neural system can be used for control of different kinds of mobile robots.

Designed by Oleg Ignatiuk
The Laboratory of Artificial Neural Networks
Brest State Technical University, Belarus