BP neural network optimization method based on improved grey wolf algorithm

Abstract

The invention discloses a BP neural network optimization method based on an improved grey wolf algorithm. The BP neural network optimization method comprises the following steps: (I) selecting the structure of a BP neural network; (II) initializing a grey wolf population by using a complex numerical coding mode, initializing parameters A, alpha and C, and determining the maximum number of iterations; (III) determining a neural network fitness function, and outputting an excitation function of a node; (IV) calculating fitness values of gray wolf individuals, finding out an optimal solution, a suboptimal solution and a third optimal solution of the fitness values, and updating position information of the rest gray wolf omega and values of parameters A, alpha and C; (V) selecting a training sample and a test sample to carry out an experiment, and recording errors and corresponding optimal solutions; (VI) judging whether the maximum number of iterations or the set error value is reached ornot; and (VII) finally returning a result of the position of the grey wolf alpha, the position of the grey wolf alpha iterated each time in the training process, the minimum error of the position ofthe grey wolf alpha, and the error of the training sample and the test sample.

Description

technical field [0001] The invention relates to the technical field of algorithm optimization, in particular to a BP neural network optimization method based on the improved gray wolf algorithm. Background technique [0002] In recent years, swarm intelligence optimization algorithm has been widely used in solving complex problems because of its simple structure and easy implementation. Inspired by the predation behavior of gray wolves, Australian scholar Seyedali Mirjalili et al. proposed a new swarm intelligence optimization algorithm in 2014: the gray wolf optimization algorithm, or GWO algorithm, by simulating the predation behavior of gray wolves and based on the mechanism of wolf group cooperation To achieve the purpose of optimization, the algorithm has the characteristics of simple structure, few parameters to be adjusted, and easy implementation. There is a convergence factor and an information feedback mechanism that can be adjusted adaptively, and it can achieve a...

AI Technical Summary

Problems solved by technology

[0003] In 2015, Long Wen et al. aimed at the problems of low solution accuracy, slow convergence speed, and poor local search ability of the basic Gray Wolf Optimization Algorithm (GWO), introduced the good point set theory to generate the initial population, and laid the foundation for the global search of the algorithm; Yang Hongguang et al. Extend the gray wolf optimization algorithm to the field of cluster analysis, and propose a new clustering algorithm (GWO-KM) that combines the gray wolf optimization algorithm and K-means

[0005] It can be seen from the above that as a new algorithm, the theory of gray wolf optimization algorithm is not yet fully mature, and the improvement and application research of gray wolf optimization algorithm are still in the initial stage, and its application research in the field of image processing has just begun

Images