Nautical bottom sediment research: Sub report 9. Statistical analysis
Wildemeersch, K.; Van Hoestenberghe, T.; Claeys, S.; De Schutter, J.; Vanlede, J.; Van Oyen, T.; Verwaest, T.; Mostaert, F. (2014). Nautical bottom sediment research: Sub report 9. Statistical analysis. Version 6.0. WL Rapporten, 00_161. Flanders Hydraulics Research: Antwerp. V, 66 + 3 p. bijl. pp.
Deel van: WL Rapporten. Waterbouwkundig Laboratorium: Antwerpen. , meer
|
| Beschikbaar in | Auteurs |
|
Documenttype: Projectrapport
|
| Trefwoorden |
Hydraulics and sediment > Hydrodynamics > Tides
Hydraulics and sediment > Sediment > Cohesive sediment
Nautical bottom
Numerical modelling
Physical modelling
Physics > Mechanics > Rheology
Sedimentation > Diagenesis > Consolidation
Sediments > Clastics > Mud
ANE, België, Brugge, Haven van Zeebrugge [Marine Regions]
|
| Author keywords |
|
ContactgegevensOpdrachtgever: Vlaamse overheid; Beleidsdomein Mobiliteit en Openbare Werken; Vlaams Ministerie van Mobiliteit en Openbare Werken; Departement Mobiliteit en Openbare Werken; Afdeling Maritieme Toegang , meer
| Auteurs | | Top |
- Wildemeersch, K.
- Van Hoestenberghe, T., meer
- Claeys, S., meer
- De Schutter, J., meer
|
|
|
| Abstract |
In this sub report the available data from the measurements in the sediment test tank (STT) is analysed with the objective to find relationships between the different parameters that are measured in the STT. A first type of analysis carried out is a correlation research. This research shows if parameters are related and in what extent knowledge about one parameter gives insight into a different parameter. A second part of the report deals with regression analysis. The main objective of this kind of analysis is to find out how well one parameter (in this scope called a variable) can be predicted by knowledge of a different set of parameters (variables). Different regression approaches are used. A first is the stepwise regression analysis. This kind of analysis tries to find a linear relation between the dependent and independent variable in such a way that it includes only those variables that are significant for the regression equation. To optimise the results obtained by the stepwise regression analysis different pre-processing techniques have been applied to the data. A first is Cluster formation, which allows to subdivide the data set in different zones (clusters) and perform a regression analysis on the subset rather than the full dataset. This approach allows more accurate predictions in the clusters when comparing to the whole dataset. A second technique is that of the Box-Cox transformation which transforms variables in the dataset in such a way that it corresponds more with a linear relationship which, again, allows better prediction. A last part of the regression analysis is nonlinear data fitting, which allows more complex (nonlinear) relationships to be fitted in order to obtain an even higher accuracy. A last part of the research studies the dynamic yield stress and attempts to find correlations between the different characteristics of the shear curves.. |
|
