Routing algorithm by using extend Q value-based dynamic programming with Boltzmann Distribution

In this paper, a method is proposed which evaluates Q Value-based Dynamic Programming with Boltzmann Distribution in static road network. The method could distribute vehicles well on all the possible routes from the origin to destination according to the probability. And in this paper, a new method is proposed to calculate the probability which considers the expected traveling time to the destination more efficiently, where two parameters τ and m are used. The proposed method is evaluated using the Kitakyushu city road network data. The simulation shows that better results could be obtained using appropriate τ and m values.