Fuzzy Measure Theory
Parameter control in evolutionary algorithms
IEEE Transactions on Evolutionary Computation
Hi-index | 0.00 |
For the actual need of future research and application, this paper proposes a new method that is a new fuzzy control system of fuzzy integral-genetic algorithm (FIGA). By fuzzy integral, it can study comprehensive evaluation of population diversity and individual quantity on three attributes: individual difference extent, the difference extent of individual's fitness and the difference extent of population lifetime, thereby dynamically adjust the rate of crossover (Pc) and mutation rate (Pm) in genetic algorithm. It improves the controller of fuzzy control for parameters Pc and Pm of genetic algorithm. The results of experiment show that the proposed genetic algorithm, combining fuzzy measure and fuzzy integral, performances better than simple genetic algorithm (SGA).