Evolving systems: Control and stability inheritance in self-assembling structures

We propose a new framework called Evolving Systems to describe the autonomous self-assembly of actively controlled dynamical subsystems into an Evolved System with a higher purpose. We introduce Evolving Systems and explore the essential topics of the control and stability properties for such systems. The Evolving Systems framework developed in this thesis provides a scalable, modular architecture to model and analyze the subsystem components, their connections to other components, and the Evolved System.;Autonomous assembly of large, complex structures in space is one application of Evolving Systems. Future space missions will entail systems where the scale, complexity, and distance preclude astronaut assistance due to the inherent risks and costs. These considerations suggest the need for a framework and methods to advance the state of the art of autonomous assembly of complex systems. A critical requirement for autonomously assembled structures is that they remain stable during and after mating. The important topic of stability in Evolving Systems is the primary focus of this research. We introduce the key component controller design approach to maintain stability in Evolving Systems. One key component from the Evolving System is augmented with additional local control to achieve the goal of ensuring system stability during assembly. The other components of the Evolving System are left unchanged, enabling modular system design and reuse of components. We present simulation results demonstrating the successful use of these methods to maintain stability in illustrative examples.;Aerospace systems are difficult and costly to model, due to their complexity and their uncertain operating environments. The adaptive key component controller we present is well suited to poorly modeled systems because its gains adapt to the sensed system outputs. We develop an impedance-admittance formulation of the contact dynamics between components of an Evolving System to obtain sufficient conditions for successful application of key component controllers. We present important results related to the inheritance of stability, passivity, and dissipativity traits of Evolving Systems.