GA optimisation of rule base in a fuzzy logic control of a solar power plant

A genetic algorithm (GA) is formulated to optimise the rule base of a fuzzy logic controller (FLC) in a solar power plant. The rule base embodies an empirical set of 49 'if-then' rules. The influence of each rule is scaled by a weighting factor which is encoded in the gene of a chromosome. The entire chromosome encodes all of the 49 weighting factors. Evaluation of the fitness of the chromosome is based on the response time of the plant. Considerably improvement of plant performance is shown after some 80 generations of evolution of the chromosome.