An estimation-based self-triggered control design for multi-agent systems over lossy channels

Simple item page
dc.contributor.advisorPesch, Dirk H J
dc.contributor.authorIjaz, Zohaiben
dc.contributor.funderScience Foundation Irelanden
dc.date.accessioned2025-05-09T13:46:21Z
dc.date.available2025-05-09T13:46:21Z
dc.date.issued2025en
dc.date.submitted2025en
dc.description.abstractMulti-agent systems (MASs) are popular for autonomous decision-making and control for a wide range of distributed systems. A multi-agent system consists of many autonomous agents that perform independent tasks but interact with their peers to achieve a common objective. MASs are extensively used in traffic control and safety systems, smart buildings, intelligent transport systems, electric power grids, industrial automation, etc. The agents in a MAS exchange information with their peer agents to jointly perform given tasks. In practice, MASs are often constrained by communication resources (limited shared bandwidth) and limited energy (battery-powered), necessitating limiting the agent's information exchange. To reduce communication and minimize energy consumption, this thesis has studied estimation-based adaptive self-triggered control (STC) techniques that limit communication between agents to key events in the systems. It also studied the challenges of communication impairments and possible solutions in designing STC for real-time systems. One of the common objectives of agents in a MAS is to develop a consensus with their peer neighbors. To develop consensus, agents need to exchange state information with their neighbors, where communication between the agents is critical. This information exchange attains more importance if the consensus objective of a MAS continuously varies with time, and the performance of the MAS deteriorates drastically if the information exchange between the agents is not perfect. To address this, a time-varying consensus of MAS over a lossy wireless channel is designed in Chapter 3, and the impact of communication losses over the MAS is studied in this work. A linear regression-based estimation mechanism is designed to estimate the lost state information due to communication impairments. One of the reasons for communication impairments is congestion over the communication channel. This is due to limited channel bandwidth or packets being sent over the channel beyond its capacity. Agents exchange information periodically with their neighbors, and the periodicity is generally kept low to account for any changes in the system. This high information exchange continuously occupies the channel, leading to congestion over the channel. Therefore, an aperiodic self-triggered control (STC)-based communication mechanism is introduced in Chapter 4 as an alternative to periodic communication, which limits the information exchange between the agents only to events in the system. The proposed STC in this thesis is studied in both synchronous and asynchronous modes. Synchronous STC is designed in Chapter 4 for centralized MAS, where a central control unit (CCU) determines the STC-based communication time. Synchronous STC is also designed for distributed MAS in this chapter, in which each agent will calculate the next STC time at the current communication instance and then exchange its state and the next STC time information with the connected neighbors. All the connected neighbors are then decided to communicate next on the minimum STC time calculated. The proposed synchronous STC is designed and validated for linear and nonlinear MASs. While synchronous STC for MAS is conceptually easier to analyze, designing this over a real-world distributed system is complex due to the real-time challenges of synchronizing all the agents. The synchronous STC can still be achieved for centralized MAS since the CCU calculates and communicates the STC time. However, this communication synchronization is challenging for distributed MAS since all the agents have to transmit at each triggering instance, possibly leading to excessive resource utilization. Therefore, an asynchronous STC is designed in Chapter 5, in which each agent will communicate at its own communication time. The challenge of communication synchronization for distributed systems is avoided in this work. The proposed asynchronous STC is designed for MASs that operate over a lossy communication channel. The communication impairments due to the lossy communication channel impact the STC design since the STC is designed using the latest state information from neighboring agents, and the accurate STC sampling time will not be calculated if the latest state information is lost due to losses. Therefore, a regression-based state estimation mechanism is designed to estimate the latest state information to overcome the challenge of losing state information due to packet loss.en
dc.description.statusNot peer revieweden
dc.description.versionAccepted Versionen
dc.format.mimetypeapplication/pdfen
dc.identifier.citationIjaz, Z. 2025. An estimation-based self-triggered control design for multi-agent systems over lossy channels. PhD Thesis, University College Cork.en
dc.identifier.endpage125en
dc.identifier.urihttps://hdl.handle.net/10468/17406
dc.language.isoenen
dc.publisherUniversity College Corken
dc.relation.projectinfo:eu-repo/grantAgreement/SFI/SFI Research Centres Programme::Phase 1/16/RC/3918/IE/Confirm Centre for Smart Manufacturing/en
dc.rights© 2025, Zohaib Ijaz.en
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/en
dc.subjectControl over lossy communication channelen
dc.titleAn estimation-based self-triggered control design for multi-agent systems over lossy channelsen
dc.typeDoctoral thesisen
dc.type.qualificationlevelDoctoralen
dc.type.qualificationnamePhD - Doctor of Philosophyen
Files
Original bundle
Now showing 1 - 2 of 2
Thumbnail Image
Name:
IjazZ_PhD2025.pdf
Size:
7.19 MB
Format:
Adobe Portable Document Format
Description:
Full Text E-thesis
Download
Thumbnail Image
Name:
IjazZ_PhD2025_Submission for examination form.pdf
Size:
608.97 KB
Format:
Adobe Portable Document Format
License bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
license.txt
Size:
5.2 KB
Format:
Item-specific license agreed upon to submission
Description:
Download
Collections
Research Theses
College of Science, Engineering and Food Science - Doctoral Theses
Computer Science - Doctoral Theses