Model predictive control for inventory management in biomass manufacturing supply chains

This paper presents a centralised model predictive control strategy applied to biomass inventory control in sugarcane industries. Sugarcane industries are important renewable energy producers and an adequate inventory control of their feed material (biomass) can improve energy production. Simple linear discrete-time models with dead-time are developed to predict the controlled variable behaviour. Two layers are used in the controller, in the upper one performance is optimised by an linear programming (LP) algorithm and a multivariable generalised predictive controller (GPC) or multivariable generalised predictive controller with dead-time compensation (DTC-GPC) is used in the lower level. Simulation results in general show that the proposed controllers globally optimise the system behaviour and find an optimal ordering amount for keeping stock levels. In cases of plant/model mismatch DTC-GPC can have a significant and positive impact on the control of stock levels adding one more parameter for achieving minimised oscillatory performances (bullwhip effect).