GA-BASED FUZZY-LOGIC CONTROL OF A SOLAR POWER-PLANT USING DISTRIBUTEDCOLLECTOR FIELDS

A novel genetic algorithm (GA) based fuzzy logic control (FLC) system has been developed for the solar power plant, Plataforma Solar de Alme ria (PSA) at Tabernas, in Almeria, Spain. The rule base encompasses an empirical set of 49 'if-then' rules. Chromosomes consisting of 49 gen es of 5-bit data are created to link to the rule base. The 5-bit data of each gene represents the strength of the corresponding 'If-Then' ru le. The GA performs the basic operations of reproduction, crossover an d mutation on a pool of chromosomes to search for the best rule base w hich optimises the response time of the plant to input temperature or power demand by controlling the distributed collector field of the pla nt. The collect field is essentially an array of parabolic mirrors and oil pipes in which the transversal of solar energy takes place. Simul ation results on the plant with an optimised rule-base using the 100th generation of the chromosome show that the proposed GA-FLC scheme giv es a better and more robust performance of the plant than other scheme s previously implemented. (C) 1998 Published by Elsevier Science Ltd. Ail rights reserved.