Menu:

Stiction Detection

Example image - aligned to the right

The Stiction Detection and Quantification package of PATS was developed to detect static friction (stiction) in control valves and to estimate the numerical values of the parameters used in the data-driven stiction models. The presence of stiction in a valve can contribute to oscillatory behaviour of a controller system. The data-driven models describe the behaviour of a sticky valve using a fixed number of parameters. This package allows the user to combine stiction detection and quantification.

The developed programme is based on nonlinear system identification. A Hammerstein model, which assumes nonlinearities are present in the input, as the basis for nonlinear system identification. The stiction model used in this programme was introduced by He et. al. (2007) and is able to accurately capture the dynamics of a faulty valve using only two parameters. The inputs to the programme are routine process data: the process output (PV) and control signal (OP), while the outputs are the estimates for stiction parameters. Allowing the user to give the programme multiple data series, downsample the data, choose different time windows, specify different linear model structures, and adjust the accuracy of estimation provide a user-friendly environment. The results of the estimation procedure are shown visually to the user.

The algorithm used by programme can be stated in the following steps:

Manual

The manual for using the toolbox can be found here.

System Requirements

This programme has the following prerequisites:

Download

The files can be downloaded from the Downloads section.

Developers

This software has been developed by the following people:

References