A New Artificial Immune Network Model for Mobile Robot Path Planning

Abstract

To solve the mobile robot path planning in uncertain environments, a new path planning algorithm is presented on the basis of the biological immune network. The environment surrounding the robot is taken as the antigen, and the behavior strategy of robot is taken as the antibody. The selection model of antibody concentration is defined based on the Jernes idiotypic immune network hypothesis, and the mobile robot path planning is realized through the selection of the antibody concentration. The simulation of path planning for mobile robot in multi-obstacle environments shows that the robot can find a safe path in complicated environments, which verifies the better adaptivity of proposed planning model. The simulation in dynamic environments shows that the robot can safely avoid all dynamic obstacles, which verifies the better flexility of new algorithm.

A real-time immune planning algorithm incorporating a specific immune mechanism for multi-robots in complex environments

“…In 1986, an immune kinetic model 11 was proposed by Farmer to calculate the stimulation value and concentration of the antibody for Jerne’s idiotypic immune network. The model can be described as…”

“…However, because the antibody and antigen are defined only according to the azimuth between the robot, obstacle and goal, and the distance between them has not been considered, the searching capability of IPCAIN is restricted. In 2013, taking the environment and robot behavior as the antigen and antibody, respectively, Yuan et al 11 presented an improved network planning algorithm by optimizing the immune kinetic model. Later, to further improve the immune network planning ability in multi-obstacle environments, a memory-based mutual-coupled immune network planning algorithm (MMCINPA) was proposed.…”