Method of Autonomous Adaptive Control Application for Group of Robots on Example of Nano-Satellite Cluster Model

Abstract

The paper presents first results of application of Autonomous Adaptive Control (AAC) method [4, 10] for adaptive control of group of robots on example of model of nano-satellite cluster. The AAC method is the biologically-inspired method, considerate to authors conceptual model of nervous system. From technical viewpoint the AAC system is self-learning recognition-control system, which has two goal functions: gnoseological one with goal of search and accumulation of knowledge and optimization one with goal of perform the optimal control of given controlled object. The AAC system contains subsystems: pattern formation and recognition, knowledge base, decision making, emotion modeling and some other subsystems. The AAC system provides self-learning and control in one process. The AAC system works on base of empirical knowledge and does not use of mathematical models of controlled object. In case of robot group the AAC system is uses as individual control system of each robot. Each individual robot learns to keep local order in the group. The global order is the result of sum of local orders. As example we have used a model of cluster of nano-satellites in experiment of monitoring of ionosphere of Earth. In the space mission the group of nano-satellites has to keep global order in the space. In our approach individual AAC systems self-learn to keep local orders in the cluster. As result we can see obtaining of global order in the cluster. The AAC system gives more cheap approach to high quality control than other methods regards to its adaptive properties.