Dynasearch Algorithms for Solving Time Dependent Traveling Salesman Problem

Time dependent traveling salesman problem(TDTSP)is an extension of the traveling salesman problem(TSP),in which the travel time or cost between two nodes depends on not only the distance between the nodes,but also the time of day or the node position in the Hamilton cycle.In this paper,we formulate the model for the TDTSP based on the node position in the Hamilton cycle,and develop the novel dynasearch algorithms to solve it.Simulation shows that the dynasearch algorithms are superior to the most effective dynamic programming heuristics in local search area.