Fuzzy Irrigation Decision Support System

Water is a limiting factor in agriculture and when improperly managed reduces the yield potential of crops. The objective of this study is to develop a Fuzzy Irrigation Decision Support System (FIDSS) to optimize water management for soybean production. Management of irrigation systems for greatest benefit requires an understanding of many physical, biological and chemical processes, and economical factors. Such processes are very complicated and involves many uncertainties. During the last 20 years, considerable progress has been made in developing computer crop growth simulation models. These models are developed with the desire to incorporate quantitatively the fundamental mechanisms controlling the above processes. However, this objective has been compromised by simplifying the mechanisms’ representations by general process descriptions which use empirical relationships determined from experimental data. In most cases, it is not possible to collect and correlate data for all conditions due to the nature of these processes. For example, variables such as soil water uptake by roots, soil water content at various depths and water lost by evapotranspiration interact with each other in a complex way making it hard to determine fundamental quantitative relationships among these variables. This study is therefore undertaken to begin the development of a mechanistic model using qualitative rather than quantitative relationships of the most important variables for making irrigation decisions. The Fuzzy set theory introduce by Lofti Zadeh in 1965 provides the framework to cope with ambiguity and unnaturalness of the traditional crisp method in this domain. Fuzzy expert systems utilize human experience and decision methods, and they have proven to be valuable when dealing with non-linear and complex relationships. In FIDSS we have developed membership functions for 21 important variables of three primary component s: 1) variables for plant, such as, leaf area index (LAI) and evapotranspiration (EP); 2) variables for water status and flow in soil, such as, soil-water content in a layer (0~) and infilteration (IN); and 3) variables for weather, such as, maximum temperature (TM) and solar radiation (RAD). FIDSS has 163 rules in its knowledge base specified as follows: