Economic optimisation of cold production: a matheuristic with artificial neural network approach
Résumé
In this paper, the economic optimisation of cold stores with a dynamic electricity price is studied. The main cost in cold storage is energy, and the combination of three factors make the optimisation of cold-room costs difficult to achieve. First, the temperature inside the cold rooms must be maintained within the allowed range, and complex thermodynamic processes make predicting temperatures a hard task. Second, production schedules that induce higher risks of compressor wear, and thus high maintenance costs, have to be avoided. Third, the price of electricity, which is known in advance, varies during the day. In this paper a cold store model is presented, which includes a fixed cost payment when the compressor production level changes, as well as an Artificial Neural Network model for the temperature. A Dynamic Programming solution is presented to solve simple cases and a matheuristic algorithm is designed to solve the general case. The matheuristic is compared to a linear solver, and performs better overall. Finally, managerial insights are derived from this study. The impact of storage capacity on operating costs is evaluated, and the influence of the fixed cost value on the electricity cost value of the optimised production schedules is discussed.
Fichier principal
Optimization_of_cold_production_with_time_varying_energy_cost__a_matheuristic_with_neural_network__IJPR_.pdf (994.43 Ko)
Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)