Author ORCID Identifier
Semester
Summer
Date of Graduation
2023
Document Type
Thesis
Degree Type
MS
College
Statler College of Engineering and Mineral Resources
Department
Mechanical and Aerospace Engineering
Committee Chair
Dimas Abreu Archanjo Dutra
Committee Co-Chair
Guilherme Pereira
Committee Member
Jason Gross
Abstract
Swarms are groups of robots that can coordinate, cooperate, and communicate to achieve tasks that may be impossible for a single robot. These systems exhibit complex dynamical behavior, similar to those observed in physics, neuroscience, finance, biology, social and communication networks, etc. For instance, in Biology, schools of fish, swarm of bacteria, colony of termites exhibit flocking behavior to achieve simple and complex tasks. Modeling the dynamics of flocking in animals is challenging as we usually do not have full knowledge of the dynamics of the system and how individual agent interact. The environment of swarms is also very noisy and chaotic. We usually only can observe the individual trajectories of the agents. This work presents a technique to learn how to discover and understand the underlying governing dynamics of these systems and how they interact from observation data alone using variational inference in an unsupervised manner. This is done by modeling the observed system dynamics as graphs and reconstructing the dynamics using variational autoencoders through multiple message passing operations in the encoder and decoder. By achieving this, we can apply our understanding of the complex behavior of swarm of animals to robotic systems to imitate flocking behavior of animals and perform decentralized control of robotic swarms. The approach relies on data-driven model discovery to learn local decentralized controllers that mimic the motion constraints and policies of animal flocks. To verify and validate this technique, experiments were done on observations from schools of fish and synthetic data from boids model.
Recommended Citation
Jimoh, Hafeez Olafisayo, "Imitation Learning for Swarm Control using Variational Inference" (2023). Graduate Theses, Dissertations, and Problem Reports. 12115.
https://researchrepository.wvu.edu/etd/12115
Included in
Acoustics, Dynamics, and Controls Commons, Controls and Control Theory Commons, Hardware Systems Commons, Robotics Commons