Graduation Year

2024

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Mechanical Engineering

Major Professor

Tansel Yucelen, Ph.D.

Committee Member

Kyle Reed, Ph.D.

Committee Member

Rajiv Dubey, Ph.D.

Committee Member

Yasin Yilmaz, Ph.D.

Committee Member

Kevin Moore, Ph.D.

Keywords

Control with local information exchange, Control under limited resources, Stability analysis, Control applications

Abstract

The overarching objective of this dissertation is to introduce system theoretical frameworks in control of misbehaving multiagent networks subject to limited resources (i.e., limited number of driver nodes), and the presence of disturbances and/or uncertainties (i.e., misbehaving nodes). Below, we briefly summarize each problem studied in this dissertation with their proposed solutions.

Specifically, we first consider multiagent networks subject to misbehaving nodes over fixed undirected graph topologies. To address this problem, we apply proportional-integral control signals to a subset of nodes (i.e., driver nodes) in the network to address the undesired effects arising from misbehaving nodes. Since multiagent networks with undirected graph topologies have limited applications in real-life scenarios, we next generalize this result to directed graph topologies. In this generalization, we not only present stability analysis but also introduce a graph-theoretical tool to provide a practical insight to users how to select the driver nodes in the network.

Alongside demonstrating the effectiveness of our proposed control approach to suppress the adverse effects of misbehaving nodes, we also consider the potential downsides of our solution. Our investigations reveal additional challenges, which lead us to explore alternative strategies to address these issues. Specifically, we first consider that the proposed control approach relies on the state information of the driver nodes in the multiagent network to generate the control signal for the driver nodes. To relax this requirement, we utilize other nodes called “observer nodes" in the network, which provide their state information to the driver nodes. Moreover, we investigate how to maintain the privacy of the global network objective (i.e., the synchronization command) while suppressing the undesired effects of misbehaving nodes. To address this potential problem, we select a certain node called “leader node", which is the only node in the network having the knowledge of the desired command. In addition, we use control nodes (i.e., driver nodes) and auxiliary sensing nodes (i.e., observer nodes). In our other study, we reinvestigate one of our previous researches, which focuses on controlling multiagent networks subject to a misbehaving node in a continuous-time setting. Our major contribution is to provide the proportional-integral controller in a discrete-time setting and demonstrate the stability analysis. Finally, we present an application of the proposed proportional-integral control in an experimental setup, which is composed of Crazyflie 2.0 nanoquadcopters and loco positioning system. In particular, we investigate the response of quadcopters when one of them is subject to disturbance (i.e., wind effect or tied to the ground). Our control algorithm ensures the stability of the overall multiagent network while suppressing the undesired effect of the misbehaving quadcopter.

The stability and performance characteristics of all of the aforementioned control algorithm are rigorously given using system-theoretical methods. Their efficacy is further demonstrated through illustrative numerical examples and/or an experimental environment.

Download

Share

COinS