Skip to Main content Skip to Navigation

Delay estimation and predictor-based control of time-delay systems with a class of various delays

Abstract : Time-delay is a widely-found phenomenon (i.e. physical dead time, communication latency, computation time) in real control systems, which can degrade the performances of the system or destabilize the system. If the time-delay is small, then the closed-loop stability can be guaranteed with conventional control techniques; but these techniques are no longer effective if the time-delay is long. This thesis is dedicated to the control of time-delaysystemswithunknown or uncertain long time-delays. In order to compensate long time-delays, the predictor-based control technique is adopted, and the delay estimation techniques are developed to assist the predictor-based controller. According to the different types of the systems and the time-delays, three objectives are analyzed in the thesis. The first objective considers the control of LTI systems with unknown constant delays, a new type of delay estimator is proposed to estimate the unknown time-delays, then it is plugged into apredictor-based controller to stabilize the system. The second objective focuses on the practical stabilization of remote control systems with unknown time-varying delays, at this time, the time-delays are estimated by a practical way: a specific communication loop is used to estimate the round-trip delay in finite time, and the system is stabilized with a predictor-based controller. This practical delay estimation algorithm is implemented on a real WiFi network, it can estimate the time-varying delays with good performances and robustness. The last objective is devoted to the control of networked control systems with time-varying delays, the discrete predictor-based control techniques are used to compensate long time-varyingdelays,and the packet reordering in the sensor-to-controller channel is also considered. Moreover, this control solution is validated on a networked visual servo inverted pendulum system, and the control performances are fairly better than the non-predictive control methods.
Document type :
Complete list of metadata
Contributor : Abes Star :  Contact
Submitted on : Thursday, March 18, 2021 - 3:16:11 PM
Last modification on : Wednesday, April 27, 2022 - 4:16:31 AM
Long-term archiving on: : Monday, June 21, 2021 - 9:02:31 AM


Version validated by the jury (STAR)


  • HAL Id : tel-03173524, version 1


Yang Deng. Delay estimation and predictor-based control of time-delay systems with a class of various delays. Automatic Control Engineering. École centrale de Nantes, 2020. English. ⟨NNT : 2020ECDN0014⟩. ⟨tel-03173524⟩



Record views


Files downloads