| Hopfield neural network is widely used in associative memory,pattern recognition,optimization calculation and optical equipment parallel implementation.The continuous Hopfield neural network with optimization capability is the focus of the research.Hopfield neural network algorithm is a classic algorithm to solve the problem of travel agents,but the traditional Hopfield neural network has the disadvantages of poor robustness and slow convergence.In recent years,researchers have proposed fractional Hopfield neural network.In this paper,the fractional Hopfield neural network optimization algorithm and the dynamic step based fractional neural network algorithm are proposed for continuous Hopfield neural network,and the convergence is proved.In addition,the rationality and effectiveness of the algorithm are verified by numerical experiments.The main work of this paper is as follows:1.A fractional Hopfield neural network structure optimization learning model is proposed.In this algorithm,the original updating formula is improved to the fractional updating formula,so as to solve the traveling salesman problem.2.Numerical experiments show that the proposed algorithm is more effective than the traditional Hopfield neural network algorithm.3.A fractional order Hopfield neural network structure optimization learning model based on dynamic step size is proposed.In this algorithm,the traditional step size is replaced by the dynamic step size,which overcomes the shortcomings of the first method,such as oscillation and less effective paths.4.The rationality of the algorithm is verified by numerical experiments,and the advantages of the algorithm are illustrated by comparative experiments.5.The Caputo Fabrizio fractional Hopfield neural network optimization learning model is proposed,and numerical experiments are carried out to select different city scale traveling salesman problems.The results of different algorithms are compared to verify the rationality and effectiveness of the algorithm. |
PDF Full Text Request