Using Fuzzy Adaptive Genetic Algorithm for Function Optimization

The most challenging problem of traditional genetic algorithms is how to achieve optimal accuracy in acceptable time. The key to improvements are suitable mutation and crossover rates. In this paper, an improved genetic algorithm, called fuzzy adaptive genetic algorithm (FAGA), is proposed. The enhanced algorithm dynamically adjusts its mutation and crossover rates according to a fuzzy inference model and the performances of individuals and populations. The proposed algorithm incorporates an elitism strategy to conserve good solutions. In addition, new individuals are introduced to guarantee population diversity and to extend the search space of the problem. The proposed algorithm is applied to several function optimization problems. The simulation results show that the average performance of the proposed algorithm overall is better than the best results obtained using a traditional elitist-based genetic algorithm