Barge Planning and Routing with Uncertainty of Container Availability

Transportation of containers in hinterland connections is preferably done by barge. This transport depends on a lot of parameters and variables. Barge planners usually try to find a solution in a continuously changing environment. This research tries to model the choices of barges planners in such a way that the model can be used in a Decision Support System (DSS) for planners. The research assumes a complete control of resources, such as barges. The demand is assumed to be stochastic, while the demand to resource allocation is also controlled. The problem is modelled as a Markov Decision Process(MDP), with notation of Rivera and Mes (2016) and solved with an Approximate Dynamic Programming (ADP) solution method as policy iteration alogrithm with basic functions. We assume that a state of the MDP can be described by some features, which can well estimate the cost of a certain state. The model is tested with a case study of a network in The Netherlands, with different initialisations and assumptions. Concluding, the model can be used to predict costs of certain actions and can be used to give an advice on certain choices.