A Hopfield-Like Neural Network in the Simulation of Traffic Flows in a Transportation Network

The possibility to use a suitable technique for dynamically simulating traffic flows in an urban transportation network is fundamental in order to operate with traffic control policies in real time. In this paper the problem of simulating traffic flows in an urban transportation network has been resolved through a particular kind of non supervised neural network. The results obtained are very satisfactory if they are compared to the ones obtained by using the conventional techniques usually used in the transportation field both in terms of accuracy and computation quickness.