Abstract
A production rule table may be defined as an ordinary table where the fuzzy levels for two conditional (if) parts of production rules are placed in the X and Y-axes and the corresponding fuzzy levels of inference (then) part are placed inside the table. Determination of the fuzzy levels of the inference (then) part corresponding to the fuzzy levels of the conditional (if) part is very complicated if it is done by simple trial and error method. Hence, this study refers to develop a method for optimization of production rule table and to use it in a fuzzy inference model. This optimization method involves four parameters to express the nature of the curvature presented by production rule table in a three-dimensional plane to make agreement between observed and calculated values. This optimization technique was applied in the model developed by Hasan et al. to determine the intake-flow size of irrigation water of headworks. Optimization of production rule table was found to be very easy to determine the fuzzy levels for inference (then) part of the production rule in a fuzzy inference model.
