ASABE Technical Library - Abstract
Member and Access Notice
Further Analysis And Modelling Of Soil Tare Level In The Sugarbeet Industry By Means Of Neural Networks
Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org
Citation: Paper number 023036, 2002 ASAE Annual Meeting . (doi: 10.13031/2013.9802) @2002Authors: G. Riva, E. Foppa Pedretti, G. Toscano
Keywords: Neural networks, sugarbeet harvesting, wastes
The unwanted soil collected during the harvesting of sugarbeets (soil tare) represents one
of main by-products of the sugar industry in Europe. This solid waste causes problems in terms of
production costs (soil tare could represent 15-20% of the transported product to the processing plant)
and environmental impact (around 2 Mt/y of soil waste must be disposed in landfills by 20 processing
plants in Italy).
In the last few years, various studies have stressed that the factors affecting the soil tare are multiple
and dependant on the nature of the soil, the machines utilised to handle the product and the
meteorological course.
In this study, we utilize different Neural Network predictors to power an online optimization scheme
for the harvesting season. The aim of the research is to better understand the relationship between
the soil tare in the sugarbeet industry and the meteorological course during the harvest and post
harvest period by means of conventional statistical and neural networks analysis.
The research was carried out on the basis of hundreds of data relevant to loads of beets (30 t/each
in average) processed in a sugarbeet industry operating in Italy (Ancona province) during the last few
years. The model obtained could be useful to better program all the harvesting and post harvesting
operations before the industrial processing. Simulated results show significant economical and
ecological advantages.