Nederlands
Optimization of a measurement procedure using an uncertainty model

Richard Kranenburg
Supervisor: Kees Vuik

Site of the project:
TNO Bouw en Ondergrond
Princetonplein 9
3584 CC Utrecht

start of the project: September 2009

In January 2010 the Interim Thesis has been appeared and a presentation has been given.

The Master project has been finished in May 2010 by the completion of the Masters Thesis and a final presentation has been given. For working address etc. we refer to our alumnipage.

Summary of the master project:
The description of the model uncertainty gives a quantification of the expected difference between the model and the measurements. Averaged over the year this description is in accordance with the measured differences. Since the accuracy of the measurements is high the observed difference can be attributed to uncertainties in the model.

As a practical application it is possible to give for every modelling result an interval such that the concentrations are in this interval with a certain probility, for instance a 1-sigma interval. At certain time steps the model can be outside this interval. Sometimes the size of the interval is large. In such a case the extra information of this interval is limited.

The first aim of the master thesis work is to develop a method where at each time step the uncertainty interval is as small as possible. Since the model is cheap to compute and only a small number of parameters is involved a Kalman filter is a good choice to achieve this.

The second aim it to apply the developed Kalman filter in order to answer questions about the measurement procedure. The main question is: which measurements reduces the uncertainty interval the most?

For all experiments we expect that the stochastic model should be adapted or enhanced. For this reason it is important to start as soon as possible with realistic applications.




Expected yearly averaged concentration NO2 (ug/m3) in the Rijnmond area for 2006



Expected concentration NOx (green) and 2-sigma accuracy interval (blue) for the concentration NOx for the first week in 2006 in measuring point Overschie (measurements in red). Aim of the master thesis project is to make the accuracy interval as small as possible

Contact information: Kees Vuik

Back to the home page or the Master students page of Kees Vuik