Modelling of coal trade process for the logistics enterprise and its optimisation with stochastic predictive control

Abstract

In the paper, a typical coal trade process is described and modelled, where one logistics enterprise with blending equipments lies in the core and two types of common contracts are elucidated to define constraints. A mixed-integer model is built and featured by addressing contract violation, blending operation, real-time price information and arbitrarily distributed stochastic demands. To deal with the stochastic demands, probabilistic constraints are formed. Accordingly, stochastic model predictive control strategy with both receding horizon and decreasing horizon formulations is developed to handle the probabilistic constraints and exploit the value of newest price information. By solving a series of mixed-integer linear programmes, optimal coal trade decisions for the logistics enterprise can be obtained, including procurement decision, selling decision and operational decision of the blending equipments. Thorough simulation experiments are carried out and compared with three different strategies, which interpret the effectiveness of the proposed strategy.