Astonishing dynamical behaviors in schools of fish and other social animal groups in nature have become the focus of multi-disciplinary studies in the past years. Rooted in control engineering and reaching out to biology, in this thesis, we aim to use biomimetic robotic fish teams as a powerful means to investigate the control and coordination issues for robotic multi-agent systems. First, to replicate the outstanding locomotion skills of real fish, we investigate the locomotion control of an individual robotic fish. Second, inspired by the observation that formations of synchronized fish may swim with higher energy efficiency, we design distributed control laws for formations of swimming robotic fish generating antiphase sinusoidal body waves. Third, based on inspiration from the coordination behaviors of fish schooling and other collective motions for social animals, we propose an evolutionary game model to control groups of robotic fish and further study the emergence and evolution of cooperation among them in multi-robot water polo matches. Finally, based on these biomimetically inspired control tools, we develop a multi-robotic fish setup using the evolutionary game theoretic ideas to construct a new framework to study the diversification of personalities and emergence of leadership that are critical for the completion of group tasks.
|Translated title of the contribution||Regeling en coördinatie van robotvissen|
|Qualification||Doctor of Philosophy|
|Place of Publication||[S.l.]|
|Publication status||Published - 2014|