Research Projects

The cloud-based distributed control of connected autonomous systems (CAS) e.g. connected robots, autonomous vehicles, and drone swarms, etc. is a complex and rapidly evolving field that combines functions of cloud computing, networked and distributed control, and multi-agent coordination under cyber physical scenarios. This type of control system of CAS will open the pathway to an application called cooperative control (e.g. path following, formation control etc.) in a synchronized way and creates a more scalable, efficient environment with high performance. However, the exchange and processing of information through open networks may cause cyberthreats e.g. data tampering and eavesdropping. Therefore, in order to design a reliable and safe cooperative cloud-based control, it is needed to develop new distributed observer based robust control systems while ensuring stability, scalability and achieving acceptable control performances without compromising information privacy and security. This proposal has considered the following challenges: • Network Issues: In CAS, agents must communicate and exchange information over a cloud network to reach a consensus in a distributed manner which is vulnerable to malicious attacks and privacy breaches. Moreover, the communication networks are prone to delays, packet loss, and bandwidth constraints etc. • Scalability: CAS often involve large numbers of agents, and as the number of agents increases, the complexity of coordination, communication, and control increases significantly. • Safety and Collision Avoidance: Connected autonomous agents must be able to avoid collisions with other agents and obstacles in their environment. • Dynamic Environment: The environment in which CAS operate is often dynamic and unpredictable e.g., in connected autonomous vehicles (CAVs), road conditions can change. • Unmeasured States: In large scale monitoring scenarios as in CAS, measuring all important state variables of agents may not be feasible due to physical and engineering constraints. From a joint observer and control design perspective, these challenges highlight a critical need to balance privacy, security and performance under parametric and network uncertainties and external disturbances. Therefore, this project forms a new innovative and progressive perspective on the complex networked control systems problem of CAS in two strands with mentioned challenges: 1) Design of privacy preserving distributed cloud based observer and H∞ robust control algorithm for CAS 2) Design of cloud enabled privacy preserving distributed observer-based H∞ robust control algorithm with differential privacy and anonymity for CAS while maintaining the stability, and control performance and enhancing scalability of CAS under two types of cyber-attack e.g. denial-of-service (DoS) and false data injection (FDI) attack and the challenges given in (1). Here, connected autonomous robot systems (CARS) are considered as the use case of CAS.