Conferences of University of Defence in Brno, AIM 2011

Font Size: 
SYSTEM MONITORING VIA HIDDEN MARKOV MODEL
Stanislav Vechet, Jiri Krejsa

Last modified: 2011-11-07

Abstract


Hidden Markov Models (HMM) are widely used in research areas where the time dependent data are measured (voice recognition is a typical application). Presented paper deals with HMMs which are used for learning the patterns in time-dependent data measured by ultrasonic distance sensors. This learned patterns are used for sensor monitoring during its measuring phase in real time. The on-line results from this monitoring are used in further processing for identification of inaccurate sensor measurement or sensor damage. Practical experiments were performed with autonomous robot Advee which is designed to move in highly populated areas such as shopping malls or congress centres.

Keywords


Hidden Markov Models, System monitoring, Sensor measurements

An account with this site is required in order to view papers. Click here to create an account.