Fuzzy modeling by ID3 algorithm and its application to prediction of heater outlet temperature

The authors propose a practical method for fuzzy modeling. The ID3 algorithm, used in the field of machine learning, was applied to select the effective variables in the premises of a fuzzy model and compute their boundary values. Even when the process had many variables, effective variables were chosen and their boundary values for fuzzification were computed automatically. This method was applied to a system to predict heater outlet temperature. Good results were obtained, and the system was operated with the required accuracy without adding new rules or without modifying rules.<<ETX>>