Network-induced time delay control to improve performance in networked dynamical systems.

摘要

In modern control systems, this era has been recognized as the third generation of control systems, which is called networked control systems (NCS). It combines communication and computation theory with the theoretical framework of control systems. Inserting a communication network in the feedback loops of a control system has many advantages, but it adds challenges such as network-induced delay. In this thesis, the main objective is to improve the performance of the NCS by proposing different methods to control the behavior of the network-induced time delay. In the first part of this thesis, we follow the control of network approach by introducing a hybrid scheduling protocol. Traffic-division arbitration (TDA) is the proposed algorithm that combines the advantages of the static and dynamic scheduling. The TDA protocol can be used as a general purpose scheduling protocol in different fields other than NCS. The second part is dedicated to decentralized control theory in NCS. In this thesis, we propose a framework that allows the decentralized control theory of non-networked systems to be applied to the decentralized networked control systems (DNCS). To analyze the stability of the DNCS under the use of the proposed framework, two approaches are considered. In the first approach, we treat the network effect as a perturbation to the non-networked system, while in the second, we treat it as a pure time delay. A study case and simulation results of decentralized control of a group of mobile robots formation is given to validate the proposed framework.