Hopfield Neural Network Based On Simulated Annealing Study

Hopfield neural networks has carried out the main body of a book mainly right away in the respect of being used to optimize studying, has suggested that owing to simulation anneal Hopfield network optimization algorithm, including that the Hopfield network optimization improves algorithm DHNN SA owing to being scattered and owing to continuous Hopfield network optimization algorithm CHNN SA. Have given a concrete algorithm out process and the flow chart in culture, having journeyed to discuss the problem (TSP) carrying out the Matlab6.5 platform finding the solution making use of combining with and the representative combinatorial optimization problem, algorithmic superiority composes in reply a pragmatism after writer data in carrying out numerical value calculation, using an instance has testified improvement.Method has set off from modern intelligence optimization, organic guarding has simulated the anneal and the network Hopfield optimization function being tied in wedlock, the condition that the minimum unties specifically for continuous type Hopfield neural networks is caught in part very strong and easy to the first value being dependent on, has adopt the simulation anneal and Hopfield neural networks to blend calculative method carrying out an improvement, has suggested that HNN SA blends an algorithm, has used simulation anneal to converge to optimum defect of optimum overall situation emboly easy to go to make up Hopfield neural networks part, Use Hopfield neural networks to go to make up the simulation anneal time function at the same time.