Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

An Inverse Calibrator For VFSMOD-W Using The Global Multilevel Coordinate Search/ Nelder-Mead Simplex Algorithm

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan

Citation:  2007 ASAE Annual Meeting  072212.(doi:10.13031/2013.23500)
Authors:   Axel Ritter, Rafael Muñoz-Carpena, Oscar Perez-Ovilla
Keywords:   Inverse modeling, parameter optimization, vegetative filter strips

When designing vegetative filter strips (VFS) for trapping sediments, the modeler faces the challenge of identifying the appropriate model parameters for the specific application conditions. When possible, local runoff/sediment inflow/outflow through the VFS and precipitation data should be collected to support the modeling predictions. Although manual model calibration is often used, this procedure generally lacks objectivity and the outcome is linked to the expertise of the user. An automated inverse optimization procedure can be used as an objective and robust model calibration alternative. The graphical user interface of VFSMOD-W, a vegetative filter strip design system, was extended to allow for the inverse optimization of the hydraulic and sediment components of the model. A robust and efficient optimization technique, the Global Multilevel Coordinate Search algorithm in sequential combination with the local Nelder-Mead Simplex algorithm (GMCS-NMS), was selected for this purpose. This is a good alternative to other existing optimization procedures, because it is adapted for solving complex nonlinear problems accurately and efficiently, does not require powerful computing resources, and initial values of the parameters to be optimized are not needed. Several examples are presented to demonstrate the benefits of the GMCS-NMS strategy compared to manual calibration when identifying the VFSMOD-W most sensitive parameters.

(Download PDF)    (Export to EndNotes)