A Rolling Horizon Based Algorithm for Solving Integrated Airline Schedule Recovery Problem

Abstract

Airline disruption incurred huge cost for airlines and serious inconvenience for travelers. In this paper, we study the integrated airline schedule recovery problem, which considers flight recovery, aircraft recovery and crew recovery simultaneously. First we built an integer programming model which is based on traditional set partitioning model but including flight copy decision variables. Then a rolling horizon based algorithm is proposed to efficiently solve the model. Our algorithm decomposes the whole problem into smaller sub-problems by restricting swapping opportunities within each rolling period. All the flights are considered in each sub-problem to circumvent ‘myopic’ of traditional rolling horizon algorithm. Experimental results show that our method can provide competitive recovery solution in both solution quality and computation time.