Solving dynamic optimization problems with adaptive networks

This paper solves dynamic optimization problems with adaptive networks based on Hopfield networks. The dynamic optimization problem includes, as a sample, a dynamic traveling salesman problem where the intercity distance of the conventional TSP is extended to time variables. In marked contrast with deterministic networks including the Hopfield network, the adaptive network can change its states adaptively reacting to inputs from the outside. It is then demonstrated that the adaptive network produces as final states locally minimum solutions to the dynamic optimization problem. It is also expected that the adaptive network is substantiated with efficient engineering devices such as VLSI. (C) 1999 Elsevier Science B.V. All rights reserved.