Prediction of Fuel Consumption for Industry by Artificial Neural Networks
DOI:
https://doi.org/10.71701/qzjt6k73Keywords:
Artificial Neural Networks (ANN), prediction, consumption, fuelAbstract
This article presents a methodology for the prediction of fuel consumption aiming to design operational strategies for a continuous supply, reducing risks and minimizing costs. This is possible if simulations are performed with predicted data. For prediction we used artificial neural networks, a methodology based on historical consumption in a period of two years or more. It was applied to five industries with different products and production cycles. The results were very acceptable considering that, in each case, fuel consumption is linked to the demand of products or scheduled or non-programmed maintenance. Therefore, it is demonstrated that neural networks are efficient for identification of fuel consumption patterns allowing their integration with other methodologies in order to optimize logistics of industries.
Downloads
References
Gil, S., Deferrari, J., & Duperron, L. (2002). Modelo generalizado de predicción de consumos de gas natural a mediano y corto plazo. 3er. Congreso Latinoamericano y del Caribe de Gas y Electricidad, Instituto Brasileiro de Petróleo e Gás, Santa Cruz de la Sierra – Bolivia.
Gil, S.& Deferrari, J. (1999). Modelo de predicción de consumo de gas natural en la República Argentina. Petrotecnia, Revista del Instituto Argentino del Petróleo y del Gas, XL(3), 27-32.
Haidar, J. (2012). Impact of Business Regulatory Reforms on Economic Growth. Journal of the Japanese and International Economies 26(3), 285-307.
Heinamann, M. (2000). Adaptive learning of rational expectations usind neural networks. Journal of Economic Dynamics and Control, 24(5-7), 1007-1026.
Morán, A. (2012). Análisis y predicción de perfiles de consumo energético en edificios públicos mediante técnicas de minería de datos (Tesis Doctoral). Universidad de Oviedo, España.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
