| dc.description.abstract | The field of cooperative, multi-vehicle systems has witnessed a significant expansion and evolution, yielding considerable opportunities for improved autonomy, resilience, and robustness. Despite these promising developments, complex challenges persist in ensuring secure and efficient rendezvous among cooperative peers. The term "rendezvous," within the realm of cooperative control, refers to the simultaneous convergence of multi-vehicle systems to a designated target location. In missile guidance, the problem of multiple pursuer missiles achieving rendezvous with a target is termed as a salvo attack. Current methodologies often grapple with issues related to synchronization, high latency and network security, all of which can adversely impact system performance and reliability. Furthermore, traditional consensus protocols tend to fall short in mitigating threats within such complex environments, leaving the system susceptible to a range of potential attacks. For example, traditional flocking or cooperative rendezvous methods utilise a shared network of position and velocity measurements to synchronize and achieve rendezvous. Malicious agents with access to the information in this network would lead to compromised strategy and potentially interrupt the interception of the pursuers with the target. These predicaments highlight an urgent need for the development of robust and innovative solutions.
Our research aims to bridge these gaps by proposing a pioneering solution: the orchestration of multi-vehicle secure rendezvous within a finite time period through the use of a shared network time-to-go consensus protocol. The essence of this approach lies in its utilization of a finite time, time-dependent control input. For the case of the multiple pursuer salvo problem, the guidance law proposed effects same time position convergence of these multiple missiles to their target location. This unique methodology is brought to life through a variety of simulation scenarios, highlighting its potential in addressing the complexities of secure rendezvous.
The scope of the research further extends to incorporate a terminal phase guidance law designed to enhance target acquisition. This is achieved by enabling more practical position convergence through state estimation of the target vehicle. A critical comparison of this approach with the nonlinear estimators extensively used in current literature affirms the viability of the proposed framework.
Building on this, the research explores the application of this convergence framework to accommodate an array of protocols within multi-agent systems. An in-depth analysis of the decentralized, leaderless network of agents, focusing on the conditions necessary for rendezvous, further reinforces the versatility and applicability of this framework.
In pursuit of an optimal solution to the secure rendezvous problem, a novel collocation-based control optimization scheme has been proposed. This strategy exhibits significant potential to extend the rendezvous framework, enabling it to cater to higher-order, centralized, objective-based requirements. The framework developed has been shown to be effective for heterogeneous, networked teams of agents. The realization of this approach would mark a significant step forward in overcoming the challenges inherent to secure rendezvous in cooperative, multi-vehicle systems, thereby bringing us closer to the ultimate goal of enhanced system autonomy, resilience, and robustness. | |
