The Research And Applications Of Artificial Neural Network In The Traveling Salesman Problem

The traveling salesman Problem (TSP) is always one of the most Interesting topics in combinatorial memorizations. In the middle seventies,the Appearance of the complexity the theory of calculation and the development of mathematical programming have greatly Improved the advancement of combinatorial memorizations. The complexity theory of calculation has proved that the so-called problem of NP and other similar Problems are equal in calculating .That is to say, we cannot use any polynomial algorithm to solve this kind of problems. From this new discovery we find that the ability of optimization method is limited and this makes it necessary for the researches to find out better solutions.An applicative analysis of HNN in solving the TSP.The improvement of HNN.The analysis of the theoretical profess of HNN in solving the TSP Problem,the new improvement of HNN on the basis of earlyimprovement,which makes the number of neural educe from n2 to (n-l)2,and improves the architectures of neural network, also improves the efficiency. This has great significances in realizing hard ware of the neural network.Aspects of parameters adaptation, selecting the number of nodes of neurons, index of winner neurons and effect of the initial ordering of the cities, as well as the initial synaptic weights architecture of the modified SOM algorithm are discussed. The complexity of the modified SOM algorithm is analyzed. The results show that the modified SOM algorithm produced better solutions than those of the existing heuristic.