Skip to Main content Skip to Navigation

Integrated diagnosis synthesis of networked control systems with communications constraints

Abstract : Networked Control systems (NCSs) have been one of research focuses in academia and industrial applications during the last few years. The advantages of NCSs over conventional or hardwired control include reduction of system wiring and increase in maintenance and troubleshooting facilities. Because of these attractive benefits, many industrial companies and institutes have shown interest in applying networks for remote industrial control purposes and factory automation. As a result of extensive research and development, several network protocols for industrial control have been released. However, several problems arise when a control loop is closed via a communication network. No matter what network is used, some network problems (e.g. delay, packet dropout, medium access constraints) can be affected a feedback loop when it is closed via communication network. Fault diagnosis and fault-tolerant control are important issues for practical control systems, especially in safety-critical systems. The theory and application of classical approaches of fault diagnosis and fault tolerant control should be revised when dealing with NCSs. Objective of this thesis is proposing new approaches to design a fault detection and isolation (FDI) system with considering network-induced effects such as packet dropout and medium access constraints. In addition, proposed algorithms of scheduling and fault diagnosis are implemented in a mini helicopter. First, Since networked-induces effects such as packet dropout and communication constraints must be considered in FDI design, an extended model for taking into account all these limitations was used. Then a strategy to develop a set of structured residuals was proposed. It guarantees robustness to packet dropout and unknown disturbances. Regarding network access limitation, sometimes it is necessary to provide a pre-defined sequence (i.e. communication sequence) before designing FDI. It describes the instantaneous medium access status of the sensors and actuators. However, choice of a communication sequence is not trivial and that is depended on the structure of the system. In addition, in practice it is not easy to find a precise mathematical model. Proposed algorithm guarantees the generation of communication sequences which preserve some structural properties of the plant. Furthermore, this algorithm can be implemented on uncertain and large scale systems. Traditionally, applications of allocation and scheduling techniques are based on offline strategies. But under offline scheduling, performance of diagnostic system may not be guaranteed when the plant is subject to unpredictable disturbances. In addition, online scheduling needs a large computation which may not be always possible in case of embedded system. Thus, a semi-online scheduling which preserve advantages of online scheduling and prevent some limitations of offline scheduling can be considered a compromise solution. Miniature rotorcraft-based Unmanned Aerial Vehicles (UAVs) are currently subject of a lot of research due to their usefulness in situations that require unmanned or self-piloted operations. They can be classified as fast dynamic systems. So, they may be ideal benchtest for studying effects of the network on performance of closed loop control of the system. The sixth chapter of this dissertation is dedicated to implementation of a fault tolerant control strategy and FDI approach proposed in previous chapters on a quadrotor application.
Complete list of metadatas

Cited literature [173 references]  Display  Hide  Download
Contributor : Dominique Sauter <>
Submitted on : Friday, October 21, 2011 - 2:27:39 PM
Last modification on : Friday, October 23, 2020 - 8:38:03 AM
Long-term archiving on: : Thursday, November 15, 2012 - 10:16:01 AM


  • HAL Id : tel-00634526, version 1



Hossein Hashemi Nejad. Integrated diagnosis synthesis of networked control systems with communications constraints. Engineering Sciences [physics]. Université Henri Poincaré - Nancy I, 2011. English. ⟨tel-00634526⟩



Record views


Files downloads