描述:
In today’s competitive marketplace, businesses face the ongoing challenge of meeting evolving customer demands while maintaining sustainable practices. For airlines, sustainability is a critical consideration that involves optimizing resource usage. This study addresses the crew recovery problem, an essential aspect of building sustainable business models for airlines. The primary objective is to minimize costs associated with crew disruptions while considering constraints, including flight time limitations. Recovery strategies, realized through actions known as recovery actions, play a pivotal role in addressing crew disruptions. Leveraging historical data, learning-based approaches have the potential to enhance algorithms for large-scale optimization problems. They provide insights that may be overlooked through traditional methods, improving the success of the recovery process. This study presents a column generation-based solution approach for the crew recovery problem, utilizing a customized deep learning model to provide recovery actions as inputs. The methodology is applied to a major European airline company. The results indicate that the model, supported by deep learning outputs, outperforms traditional methods in terms of solution quality and efficiency.
描述:
In today’s competitive marketplace, businesses face the ongoing challenge of meeting evolving customer demands while maintaining sustainable practices. For airlines, sustainability is a critical consideration that involves optimizing resource usage. This study addresses the crew recovery problem, an essential aspect of building sustainable business models for airlines. The primary objective is to minimize costs associated with crew disruptions while considering constraints, including flight time limitations. Recovery strategies, realized through actions known as recovery actions, play a pivotal role in addressing crew disruptions. Leveraging historical data, learning-based approaches have the potential to enhance algorithms for large-scale optimization problems. They provide insights that may be overlooked through traditional methods, improving the success of the recovery process. This study presents a column generation-based solution approach for the crew recovery problem, utilizing a customized deep learning model to provide recovery actions as inputs. The methodology is applied to a major European airline company. The results indicate that the model, supported by deep learning outputs, outperforms traditional methods in terms of solution quality and efficiency.