| As the rapid expansion of air transportation and lack of the resources, irregular flights happenmore and more frequently, which have been the most popular topic of passengers and a challenge ofairlines. How to reduce irregular flights and improve passenger satisfaction has become the new studytrend and become a significant factor for airlines to organize transportation, reduce operation costsand improve corporate image.According to the operation processes and demand of passenger itinerary recovery in domesticairlines, phased recovery for passenger itinerary and integrated recovery for flights and passengeritinerary was discussed and analyzed in this paper. Firstly, according to the survey of domestic airlines,irregular flight service, passenger itinerary recovery policies and processes were summarized andanalyzed. Secondly, a space-time network was built and a multi-commodity flow model of passengeritinerary recovery was established based on aircraft recovery schedule. Danzig-Wolfe decompositionalgorithm was designed to solve the model and passenger itinerary recovery schedule can be given.Actual instance analysis shows that Danzig-Wolfe decomposition can be used to find out the optimalsolution which reduces the airlines cost and enhances the passenger satisfaction in compliance withthe request of real-time decision-making. Finally, a mixed programming model of integrated recoveryfor flights and passenger itinerary was established and a greedy simulated annealing algorithmcombined with column generation algorithm was proposed to improve solving efficiency. It is provedby a real case that the algorithm can effectively solve the irregular flight recovery problems and obtaina satisfactory solution under real-time decision-making requirements. |