Improvement Of Fruit Fly Optimization Algorithm And Its Applications

Fruit Fly Optimization Algorithm(FOA) is a new kind of global optimization algorithm based on swarm intelligence. It is derived from the drosophila foraging behavior simulation in the nature. The drosophila has a marvollous sense of smell and sight, and FOA simulates these abilities of the drosophila. Compared with some swarm intelligence optimization algorithms, FOA has the characteristics of less parameters, simple calculation and fast global search ability. But the problems of low convergence precision and easily relapsing into local optimum are existed in the basic FOA, because the FOA fails to make full use of population information. Consequently, the improved FOAs are proposed in the article.The main topic of this paper is the FOA, including the improved FOAs, using the improved FOA to optimize the expansion parameter of RBF neural network, and making RBF neural network for predicting the gasoline octane number. The main works of this paper are as follows:1)Based on the basic FOA the population center of mass and the individual optimal alternative history are added into the position updated way. In this new algorithm, parameters C1 and C2 are used to adjust the population center of mass and individual optimal alternative history, and the balance the local search and global search is reached in this way. Then the performance of the new FOA is improved. The experimental results of standard test functions show that the convergence speed and precision of the improved fruit fly optimization algorithm are improved obviously.2)Based on the basic FOA the optimal of population history and the individual optimal alternative history are added into the position updated way. The last proposed algorithm enhances both the instructiveness and the population diversity. By simulating in the standard test functions, the experimental results show that the algorithm has the advantages of global searching ability compared with three improved FOA, Particle Swarm Optimization algorithm(PSO), and Artificial Fish Swarm Algorithm(AFSA).3)The Spread of RBF is optimized by using the improved FOA, and the gasoline octane number is predicted by using the optimized RBF. The experimental results show that the optimized RBF prediction model has higher precision.