Achieving consensus in robot swarms : design and analysis of strategies for the best-of-n problem
Gabriele Valentini (Author)
This book focuses on the design and analysis of collective decision-making strategies for the best-of-n problem. After providing a formalization of the structure of the best-of-n problem supported by a comprehensive survey of the swarm robotics literature, it introduces the functioning of a collective decision-making strategy and identifies a set of mechanisms that are essential for a strategy to solve the best-of-n problem. The best-of-n problem is an abstraction that captures the frequent requirement of a robot swarm to choose one option from of a finite set when optimizing benefits and costs. The book leverages the identification of these mechanisms to develop a modular and model-driven methodology to design collective decision-making strategies and to analyze their performance at different level of abstractions. Lastly, the author provides a series of case studies in which the proposed methodology is used to design different strategies, using robot experiments to show how the designed strategies can be ported to different application scenarios
Studies in computational intelligence, v. 706, volume 706
1 online resource (xiv, 146 pages) : illustrations (some color)
9783319536095, 3319536095
973879035
Introduction
Part 1:Background and Methodology
Discrete Consensus Achievement in Artificial Systems
Modular Design of Strategies for the Best-of-n Problem
Part 2:Mathematical Modeling and Analysis
Indirect Modulation of Majority-Based Decisions
Direct Modulation of Voter-Based Decisions
Direct Modulation of Majority-Based Decisions
Part 3:Robot Experiments
A Robot Experiment in Site Selection
A Robot Experiment in Collective Perception
Part 4:Discussion and Annexes
Conclusions
Background on Markov Chains