Many-to-Many Path Planning for Emergency Material Transportation in Dynamic Environment

Abstract

The problem of emergency material transportation in dynamic environment requires to find optimal path between multiple emergency material storage nodes and distribution nodes in a changing routing environment, so as to guarantee the supply of materials within the shortest time. It corresponds to a many-to-many path planning problem in dynamic routing network. The existing static plan optimization and dynamic path optimization method are difficult to ensure the theoretical optimality of the solution in a dynamic disaster environment, and may lead to the failure of emergency material transportation. In this paper, a method of co-evolutionary path optimization is proposed and improved to resolve the many-to-many path planning problems. The ripple diffusion algorithm completes the search process in the form of a ripple diffusion relay race in a given routing environment. Furthermore, the coevolutionary path optimization method combines the ripple diffusion process with the routing environment change process. When different ripples compete with each other, the routing environment changes dynamically at the same time. Finally, the theoretical optimal solution is obtained in just a single off-line operation. The experimental results show that the coevolutionary path optimization method has advantages over the traditional method in success rate, solving time, optimality, and flexibility.