Concept of neighborhood degree and its application to switching plural optimization methods in scheduling

A concept of neighborhood degree is proposed to evaluate the quality of solutions to scheduling problems such as vehicle routing, scheduling, and dispatching problems. It is possible to apply it to the optimization process of scheduling problems in order to switch between various optimization methods by considering convergence speed and solution quality. In the experiments on TSP benchmark data, two optimization methods, i.e., tabu search and simulated annealing, are switched effectively by observing the variation of the neighborhood degree. Directions for Practical applications are also mentioned.