| 摘要 |
PROBLEM TO BE SOLVED: To properly deal with various events and to retrieve an optimum solution by producing automatically a genetic algorithm and evaluating it while changing both set value of a parameter and an operator. SOLUTION: A GA operation part 150 changes in sequence the processing points in regard to every gene locus of a chromosome that is given from a chromosome initial group generation part 110 and thereafter every gene locus of a group of chromosomes of each generated generation respectively and outputs every chromosome to an evaluation function 130 to receive the evaluation value. Meanwhile, a steady state deciding part 140 decides whether an optimum path is available for a relevant gene locus and repeats this processing to decide an optimum path. Then the part 150 outputs the optimum paths so far decided up to the current generation to an optimum solution deciding part 120 together with the path evaluation value to make the part 120 decide an optimum solution. If the optimum solution is not decided, this processing is carried out for the next generation. Thus, the optimum paths are successively decided for the gene loci of every generation, an optimum solution is decided at the part 120 among these optimum paths and the final optimum solution is decided and outputted.
|
