Adaptive Model-Free Control for Robust and Dynamic Management of Perishable Products*

This paper investigates the controller design for a robust and efficient supply strategy for logistic constrained systems with perishable goods subject to uncertainties on the demand and the process delays. Here, instead of model-based control strategies, through a continuous, local identification of the level variations and forecasting via the setting of time series, an adaptive model-free control that does not need any prior knowledge of the supply chain behaviors is developed. The proposed approach takes into accounts not only the decay of goods stored in warehouse (perishing inventories) and prevents from exceeding the facilities capacities but also ensures full demand satisfaction irrespective of the value of delay and reacting significantly faster to the consumer's demand uncertainty. Quantitative analyzes prove a strong reduction of the storage cost, increase of the profit and bullwhip effect attenuation.