Submitted Abstract
A trend in current control theory research is the focus on distributed control and estimation in networked and multi-agent systems. This work is driven by innovations such as the Internet of Things, autonomous driving in platoons, and teams of unmanned aerial vehicles. A key object of study is the network of interactions. Robustness and cost efficiency require algorithms that work with sparse networks. Technological multi-agent systems share similarities with natural systems, where large swarms exhibit emergent behaviours such as synchronization. Examples include flocks of birds or schools of fish aligning their individual movement directions. These phenomena have received considerable attention, especially in biology and physics. The models used in those fields, e.g., the Kuramoto model, does not always model the network as sparse. However, since evolution also promotes robustness and energy efficiency, we expect biological networks to be sparse. Our approach, which is based on control theory, offers a complementary perspective where the influence that the network has on a biological systems is accounted for. This project concerns the theory of emergent behaviours in a selection of complex systems and its applications in systems biology. We focus on flocking in birds, schooling in fish, and aggregation in a social amoeba. This project consists of three parts: 1) Theorems and proofs regarding the convergence of swarms; 2) Fitting models and networks to real data on synchronisation in flocks and schools; 3) An in silico study of the life cycle of the social amoeba Dictyostelium discoideum. The project is integrated with ongoing work in the System Control Group, in particular with Johan Markdahl’s research on high-dimensional Kuramoto models, the doctoral training unit CRITICS on critical transitions in complex biological systems, and the PI’s extensive work on gene regulatory network inference.This project is part of an ongoing interdisciplinary effort to generalize the Kuramoto model of synchronization to systems that evolve on nonlinear spaces. We will lead the theoretical developments of this work that relate to network synchronization and explore their applications in biology. Ultimately, these efforts will result in a better understanding of the mechanistic principles that govern complex systems in biology, which is an end to itself. This knowledge can also be transformed into blueprints for control designs in artificial life, e.g. the principles for flocking of robot birds and schooling of robot fish.