Super-Twisting Algorithm-Based Time-Varying Delay Estimation With External Signal

Abstract

This article provides an online time-varying delay estimation method by using an external signal sent along with control inputs and output measurements. Since the external signal is isolated from the system, the linearity and delay-identifiability of the system are no longer required. Moreover, the sliding mode method guarantees the finite-time convergence of the delay estimation. By comparing with the standard sliding mode method, the super-twisting algorithm reduces the chattering and provides better performances. Furthermore, the super-twisting algorithm based delay estimator is implemented on a real remote data transmission system and its performances are illustrated by experimental results.

A practical online time-varying delay estimation of remote control system based on adaptive super-twisting algorithm

“…Before stating the main results of this work, one first recalls the main result of the previous work [16].…”

“…Inspired by the work of [15], a super-twisting (STW) algorithm based TDE technique is introduced in [16] with experimental results. By comparing with [15], the following improvements can be found in [16]:…”

“…• in [15], the system clock (s(t) = t) should be used to estimate the round-trip delay, but [16] works with all monotonic signals (including system clock); • [16] is more robust than [15] with respect to the "channel inherent noise" (i.e. the noise or perturbation in the communication channel or at the receiving terminal [17]).…”

“…This work is the continuation of the paper [16], the adaptive super-twisting algorithm (ASTW) [18] is used to replace the standard super-twisting algorithm. As stated in [16,Remark 1], the estimation performance depends on the parameter tuning, unsuited parameter leads to undesirable chattering [19, p.6-8] or degrades the TDE performance. The above drawback is overcome with the adaptive supertwisting algorithm, and the parameters are self-adapted (i.e.…”

“…The above discussions are illustrated by the experimental results given in subsection IV-A. In [16], it is shown that the super-twisting algorithm based TDE method is more robust than the measurement approach [15] with respect to the channel inherent noise. With the proposed method, this advantage is preserved.…”