Previsão do coeficiente de atrito em pista de pouso e decolagem utilizando Redes Neurais Artificiais

• Ano de defesa: 2020
• Autor(a) principal: Quariguasi, José Breno Ferreira
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Aeroporto; Segurança operacional; Manutenção; Aprendizagem de máquina
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/55264

The decision-making process, according to airfield pavements maintenance and rehabilitation, especially on runways, often is made subjectively or based on technical experience of services and past works. However, this approach does not allow airport operators to rationally use the available human, material and financial resources for application in the necessaries activities to maintenance or rehabilitation of these pavements. Coefficient of friction and macrotexture are used as a criteria in decision-making related to maintenance and rehabilitation services, because they are fundamental to ensure the safety during the landing and takeoff operations on runways. These variables are influenced by some factors, for example by the aircraft traffic due to the rubber accumulated on the runway from airplane’s tire, relative humidity and the age of the surface layer, among others. In this way, this work proposes a prediction model for runway friction in order to contribute to the decision-making process related to maintenance and rehabilitation measures in Airport Pavement Management System. This model was developed using Artificial Neural Network, with data from reports of coefficient of friction measured between 2015 and 2019 at the International Airport of Fortaleza, as well as climatological data from the Brazilian Airspace Control Institute. Although, it needs future improvements in order to reduce the error, the results of the developed model were satisfactory, then the following contributions can be highlighted: (i) allowing airport operators to plan maintenance and rehabilitation actions and/or field measurements, and, thus, avoiding problems or constrains related to takeoff and landing operation safety due to the surface condition; (ii) and the findings of this work may also allow the regulatory agency to inspect and/or monitor the coefficient of friction at airfield pavements.