Algorithms for change detection based on finite sample system identification theory

Abstract

Detection of abrupt changes has many important applications. One particular application that will be investigated is leak detection in irrigation channels. As agriculture accounts for about 80% of Australia's water usage, the timely detection of leaks means that corrective actions can be taken early which will lead to large water savings and significant environmental benefits. The developed methods can be designed with any false alarm rate. This is important since frequent false alarms lead to wasted resources and operators will stop using the system. The technology once developed can be transferred to many other application areas such as urban water supplies, pipelines for oil and gas, and ..

Related publications (3)

Asymptotic properties of SPS confidence regions

Erik Weyer, Marco Claudio Campi, Balazs Csanad Csaji

2017-08-01

Closed-loop applicability of the Sign-Perturbed Sums method

BC Csaji, E Weyer

2015-02-08

Sign-Perturbed Sums: A New System Identification Approach for Constructing Exact Non-Asymptotic Confidence Regions in Linear Regression Models

Balazs Csanad Csaji, Marco Claudio Campi, Erik Weyer

2015-01-01