Resilient Distributed Optimization with Trust Opinion Dynamics for Cyber-Physical Systems Over Directed Graphs

Abstract

Dear Editor, This letter deals with distributed optimization problem in multi-agent cyber-physical systems over directed graphs and the presence of malicious agents. Empowered by the advancement of communication, sensing and computation, cyber-physical systems have attracted a great deal of attention, especially in various smart manufacturing scenarios in the context of Industry 5.0 [1]. With the strong reliance of cyber-physical systems on distributed optimization methods for economic and safe functioning, it is critical to address the vulnerability of optimization methods [2]. In particular, malicious agents can hinder the normal operation of optimization schemes, driving the system to a non-optimal state or even preventing convergence. Hence, enhancing the resilience of distributed optimization has been a significant research frontier.