A New IntelliSense Strategy Based on Artificial Immune System for Multi-Robot Cooperation

2018 ◽  
Vol 30 (1) ◽  
pp. 128-137 ◽  
Author(s):  
Tao Xu ◽  
◽  
Zengyong Shi ◽  
Xiaomin Li

In this paper, a novel intelliSense strategy based on an artificial immune system for multi-robot cooperation (MRC) is proposed. A laser range finder and camera are mounted on the robot to provide environmental information. Based on the principle of the artificial immune system, environmental information sensed by the robot is considered as an antigen while the robot is regarded as a B-cell and a possible node as an antibody. To improve exploration efficiency, the immune system is utilized in the single robot system (SRS) and multi-robotic system (MRS). Antibody-antigen affinity is calculated to choose optimal number of possible node, ensuring that the exploration path is optimal. Each robot explores independently according to affinity information, and cooperation among robots can be realized by utilizing the interactions among antibodies. In order to make the map merging strategy more persuasive, the Scale-Invariant Feature Transform (SIFT) feature is further used to identify the same region. In a real-context experiment, the proposed algorithm results in more accurate exploration, and it is verified that the proposed intelliSense strategy improves the accuracy and execution efficiency for mobile robots.

Author(s):  
Muhammad Tahir Khan ◽  
Clarence de Silva

This paper investigates multi-robot coordination for the deployment of autonomous mobile robots in order to carry out a specific task. A key to utilizing of the full potential of cooperative multi-robot systems is effective and efficient multi-robot coordination. The paper presents a novel method of multi-robot coordination based on an Artificial Immune System. The developed approach relies on Jern’s Immune Network Theory, which concerns how an antibody stimulates or suppresses another antibody and recognizes non-self antigens. In the present work, the robots are analogous to antibodies and the robotic task is analogous to an antigen in a biological immune system. Furthermore, stimulation and suppression in an immune system correspond to communication among robots. The artificial immune system will select the appropriate number of antibodies autonomously to eliminate the antigens. The developed method of multirobot coordination is verified by computer simulation.


Author(s):  
Muhammad Tahir Khan ◽  
Toar Imanuel ◽  
Yelnil Gabo ◽  
C. W. de Silva

The human immune system is a network of cells, tissues, and other organs that defend the body against foreign invaders called antigens. Jerne’s Idiotypic network theory concerns how an antibody in the immune system stimulates or suppresses another antibody and recognizes an antigen. Based on the principles of the human immune system and Jerne’s idiotypic network theory this paper presents a method for cooperation among robots in a multi-robot system. The developed cooperative multi-robot system is fully autonomous and distributed. In the present paper, cooperation is not assumed a priori. If a robot is unable to complete a task alone, the system autonomously chooses the appropriate number of suitable and most capable robots in the fleet to cooperate with each other in carrying out a global task. The approach developed in the paper incorporates robustness and fault tolerance in immune system–based multi-robot cooperation.


Sign in / Sign up

Export Citation Format

Share Document