A Fault Tolerance Scheme based on Packet-loss Fault Prediction

A predictive fault tolerance scheme is proposed to predict and avoid the packet-loss fault caused by the time-varying network induced delay, and the probability that a packet-loss occurs in the future is discussed quantificationally. Firstly, the series of network induced delays are modeled as Auto-...

Full description

Saved in:
Bibliographic Details
Main Authors: Wang, Zhan-Yu, Liu, Guo-Ping, Huang, Xian-Lin
Format: Conference Proceeding
Language:English
Subjects:
Delays
Fault tolerance
Networked Control Systems
Packet loss
Prediction of Packet-loss Fault
Predictive models
Stability analysis
Switched systems
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A predictive fault tolerance scheme is proposed to predict and avoid the packet-loss fault caused by the time-varying network induced delay, and the probability that a packet-loss occurs in the future is discussed quantificationally. Firstly, the series of network induced delays are modeled as Auto-Regressive and Moving Average (ARMA) model, and the networked control systems(NCSs) is modeled as the switched systems with packet losses and the constant system delay, which is the threshold to judge whether a packet is dropped. Then, the probability that network induced delay may exceed the constant system delay at a certain-step future is given for both direct and indirect transmission cases. Thereby, the probability that a packet-loss fault occurs is obtained to evaluate the performance of the current control signal path, and when the probability is greater than a certain level, the redundancy controller and communication path in network are adopted to maintain the system stable in order to ensure the reliability of NCSs. The simulation illustrates the effectiveness and the validity of the proposed theorems and approach.
ISSN:2161-2927
DOI:10.23919/ChiCC.2018.8483479