Simulation of stock control policies in a two-stage production system application to pig slurry management involving multiple farms

One solution to the problem of surplus animal wastes in areas with a high density of industrial livestock farms is to install collectively managed treatment plants serving several dozen farms. The way these plants should be supplied by farms must be carefully planned so that any overflow of stock is avoided and a regular supply to the unit is guaranteed. Thus, for each farm one must consider two questions: "When should a delivery be made?" and "How much should be delivered?". To achieve these management goals, the model presented seeks to simulate various supply policies and optimise control parameters. The approach is based on hybrid dynamical systems, comprising both continuous and discrete variables, the former representing changing stock levels and the latter being decision or control variables. The dynamic simulation approach used to control multiple stock in such a two-stage production system is compared with classical inventory management systems, based on static optimisation models. (Résumé d'auteur