Designing Robotic Groups Under Incomplete Information About the Dominance of Many Goals

Abstract

The article presents an approach to solving the problems of choosing the best designs of robots and their combinations in the presence of incomplete information on potential goals. Such information is modeled using binary relations of preference (dominance) on the set of robots and goals. An analytical review of various definitions of optimality and dominance of robots is given. Based on the proposed models, a number of typical problems of designing groups of robots are formulated and recommendations are given for their solution. The advantage of our approach is that it allows you to find the best solutions for maximizing the achievement of many goals with limited resources. This is achieved through the use of all available information about potential goals and preferences when choosing robots designs, building them into a group that covers as many goals as possible, and choosing an appropriate strategy. The ideas of the methods are illustrated by examples.