A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering, School of Environment, Resources and Development

Thesis (M.Eng.) - Asian Institute of Technology, 1989

The advances in micro-computer technology has made expert systems receive much attention in research, professional work and many fields of technology. In this study, an expert system 11ESIM11 (Expert System For Irrigation Management) has been developed for making decision on water management in an irrigation project. A general strategy for development and application of expert systems has been proposed. Since an expert system shell 11EXSYS11 has been used as a development tool, the study focuses on knowledge engineering which mainly involves knowledge acquisition, system design, implementation and evaluation. In an irrigation management problem, expert decisions are to be taken at various levels for assessment of water availability and requirements, in proposing a type of water scheduling and on deriving an actual operational policy for various crop scenarios. The work also involves developing data bases, and external computer programs for numerical processing. Based on the statistical analysis of streamflow, a new type of index called 11 Probabilistic Scheduling Index 11 has been proposed in this thesis. The expert decision on a particular type of scheduling (Fixed/ Arranged/Demand) is based on this index together with other site specific conditions. The software developed is interactive and has been made user friendly. The expert system "ESIM" has been evaluated by applying it to the irrigation management problem of Mae-Taeng Irrigation Project situated in northern Thailand. The output of the expert system has been found to be reasonable and practically useful results were obtained so as to aid in improved water management decision making at the project.