Estimating the immeasurable: soil properties
Type:
Presented during:
CEDA Dredging Days 2009 - Dredging Tools for the Future, Rotterdam
Authors:
Braaksma J, Osnabrugge J and de Keizer C - IHC Systems, The Netherlands
Abstract: The dredging process is characterized by the strong influence of soil properties, which vary with changing excavation locations. For optimizing the dredging production and energy consumption, it is necessary to know these properties. The problem is that a lot of these soil properties are not measured or very difficult to measure. By taking soil samples, it is possible to give a raw indication, but this does not cover the total dredging area.
Equipment for measuring the properties on-line are complex and, if available at all, too expensive to use. As an alternative, we use estimation methods. These estimation methods are based on knowledge obtained with the development of training simulators in the last decade. For these training simulators, sophisticated soil models, together with the dredging equipment, are modelled to simulate the dredging process dynamically. By using specific soil parameters, it is possible to forecast the dredging dynamics accurately. In this paper we present a, for the dredging industry, novel approach which uses the models in an inversed way: using the models and the measurement data to estimate the soil type dependent parameters. To obtain this objective several advanced filters have been successfully implemented and tested in practice. Examples are recursive least squares filtering, linear Kalman filters and more complex techniques such as the extended Kalman filter and the Particle filter. In this paper four estimation examples will be described and the advantage for controlling the process control such as:
- estimating the mean grain size of the dredged soil
- estimating the overflow losses
- estimating the dredging forces
- estimating the anchor positions
The use in practice of the immeasurable will be described as well as future developments.
Keywords: Extended Kalman Filter, Particle Filter, Overflow Losses, Anchor Position, Dynamic Positioning and Dynamic Tracking, Trail Speed Control