Economic optimisation of cold production: a matheuristic with artificial neural network approach

Abstract

In this paper, the economic optimisation of cold storage is studied. A modern cold room is mainly composed of compressors, a tank to store heat-transfer fluid and cold rooms. The main cost is incurred by energy consumption and maintenance. The price of electricity, which is known in advance, varies during the day. Production schedules that entail higher risks of compressor wear, and thus high maintenance costs, have to be avoided. The temperature inside the cold rooms must be maintained within the allowed range, and complex thermodynamic processes make it difficult to predict temperature. The tank has a limited capacity. This paper presents the first model optimising the management cost of a cold store with a tank. Maintenance costs are considered for compressors, and Artificial Neural Networks are used to forecast the temperatures in the cold rooms. An optimal Dynamic Programme is designed for the case with one cold room and a matheuristic algorithm is presented for the general case with several cold rooms. A comparison with a classical hysteresis controller shows significant savings. The impact of storage capacity on operating costs is evaluated, after which the influence of the maintenance cost value is discussed.

Keywords:

• Cold storage
• energy management
• neural network
• cost optimisation
• matheuristic

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 Improving Cold storage Equipment in Europe, https://ec.europa.eu/energy/intelligent/projects/en/projects/ice-e.

2 EPEX SPOT SE, https://www.epexspot.com.

Additional information

Notes on contributors

Alnour Ribault

Dr. Alnour Ribault is a Computer Science engineer, currently working for Energy Pool on the optimisation of complex industrial processes. He obtained his PhD from Universitè Lyon 2. His main research topics are applications of operations research and data science techniques to the management of the supply chain and energy systems.

Samuel Vercraene

Dr. Samuel Vercraene, is an associate professo at INSA LYON since 2013. His main research topics are operation research and operation management. He has published several papers on reverse logistics and transportation of peoples.

Sébastien Henry

Dr. Sébastien Henry, associate professor since 2006 at University of Lyon 1, is currently head of mechanical department of IUT Lyon 1 and co-leader of the ”Information System and Data” research team of DISP Lab. His main research topics are data management, process assessment and decision making based on machine learning and model-based approaches in the fields of energy, food, mechanics, etc.

Yacine Ouzrout

Dr. Yacine Ouzrout is a Computer Scientist in the Supply Chain & PLM group of the DISP Laboratory at the University Lumiere Lyon 2. He obtained his PhD from the INSA Lyon. Currently, he is Professor and Director of the Institute of Technology of his University. His research interests include multi-agent systems, artificial intelligence, optimisation and decision support systems. He has been involved in several European projects (Erasmus+, H2020...) and he is co-chair and member of program committees & reviewer of several International conferences and Journals.