Application of genetic algorithm in extraction of fuzzy rules for a boiler system identifier

Performance of a fuzzy system identifier is investigated against a fossil fuel boiler data. A multi-layer neuro-fuzzy system presents identification of a drum type boiler. This identification technique provides a rule-based approach to express the boiler dynamics in fuzzy rules that are generated from the experimental boiler data. The interconnections of neuro-fuzzy layers furnish these fuzzy rules. A genetic algorithm (GA) trains the neuro-fuzzy identifier and extracts the linguistic rules from measured boiler data GA training uses nonbinary alphabet and compound chromosomes to train the multi-input multi-output (MIMO) neuro-fuzzy identifier. The fuzzy membership functions are tuned during the training to minimize the identifier response error. Hence, the fuzzy rule set and tuned membership functions provide identification of the boiler. Error back-propagation training methodology is chosen to tune the membership function parameters. This neuro-fuzzy identifier obtains transient response comparable to the mathematical boiler model. The identifier response is examined in several operating points of the boiler. The identification is implemented within an object oriented programming (OOP) tool that provides portability of the identification process. Therefore, the identifier program is highly structural and transferable to different plants