A Fuzzy Logic Yield Simulator For Prescription Farming

Abstract

Interest in prescription farming has grown as the technology necessary for its implementation has become available. The central concept of prescription farming is that materials (chemicals, fertilizers, seeds) are optimally applied as a function of position within the field. Therefore, profits are maximized and potential adverse environmental effects are minimized. Our objective was to describe how fuzzy logic could be used to develop a crop yield simulator for assessing spatial variability with sufficient accuracy for optimizing application rates. The method is based on predictive yield models developed using field-scale research techniques. Two conceptual, expert system models were developed using fuzzy logic rules. In one model, chemical and physical characteristics of the soil were measured and combined with local meteorological data as input parameters. In the other model, soil properties were estimated rather than measured. The fuzzy logic rule sets were implemented using a spreadsheet. Rule sets were developed to simulate yields for two 16-ha fields in central Iowa. Predicted yields were then compared with measured yields for those fields. Our results indicate that on a relative basis, predicted yields generally agreed with measured yields.