Improved Fruit Fly Optimization Algorithm Based on Simulated Annealing in Neural Network

Aiming at the problem that the fruit fly optimization algorithm (FOA) has weak local optimization ability and difficulty in convergence, this paper combines the improved FOA (IFOA) and simulated annealing algorithm (SA) to propose a new optimization algorithm called IFOA-SA algorithm. The IFOA-SA algorithm first modifies the optimization step size of FOA, increases the escape coefficient, and improves the generation mechanism of candidate solutions. At the same time, the concept of three-dimensional search is introduced to expand the search range of fruit flies, and then the FOA solution acceptance mechanism is modified through SA. Finally, the fusion algorithm IFOA-SA is used to optimize the Generalized Regression Neural Network (GRNN) and Back Propagation (BP) Neural Network for freight volume forecasting and nonlinear function fitting. The experimental results show that the total error of the BP neural network optimized by the IFOA-SA algorithm is 37.4903, which achieves better optimization results than other algorithms. In the end, the effectiveness of the algorithm is proved through experiments.