Ground Handling Process Optimization Model Linked to Flight Delay Prediction Results

Abstract

In response to the challenge of mitigating flight delays, this study introduces an innovative solution that encompasses the prediction of delay durations for existing flights and the subsequent optimization of ground service processes based on these predictions. The indirect forecasting of flight delays is achieved through the construction of a random forest model, exhibiting a remarkable 100% accuracy when considering a 15-minute standard for flight delays. In light of the delay prediction outcomes, distinct delay coefficients are assigned to individual flights, facilitating the development of a ground service optimization model through the application of a genetic algorithm. Within the genetic algorithm optimization framework, significant enhancements have been implemented in the gene encoding of the initial population, incorporating a segmented encoding approach. Employing this refined model to optimize the service sequence and duration of ground service vehicles for all flights culminates in the notable accomplishment of achieving zero delays for the entire set of flights.

Keywords

• Flight delay prediction
• ground service optimization
• random forest model
• genetic algorithm model