Aiming at the dynamic vehicle routing problem of dangerous goods under a dynamic environment, a mathematical model is established, which is to minimize the risk cost and distribution mileage in the road segment and to maximize the freight load factor. Here a cloud-based adaptive ant colony algorithm is proposed. There are crossover and mutation operations in ant colony algorithm that may lead to premature convergence and loss of late diversity. With the characteristics of dangerous goods logistics, cloud computing is introduced to design cloud crossover and mutation operators to operate this algorithm in response to the above gap. On this basis, the proposed algorithm can be improved in the context that the simulation schedule example will reduce premature probability and enhance the iterative search efficiency more than other algorithms.
