Abstract: We consider a dynamic airline ticket pricing problem aimed at maximizing revenue given demand forecasts for each booking class and limited flight capacity. Four combinatorial problem formulations are developed and explored: 1) DTA (Discriminatory Ticket Allocation): individual prices are assigned to each customer; a simple greedy algorithm is proposed, providing an upper bound on achievable revenue. 2) PDA (Price Daily Assignment): a single price is set per day; the problem is solved both via dynamic programming and exact optimization using ILP solvers, yielding a lower bound. 3) IDA (Inventory Daily Assignment): on each day, the model defines the quantity of offered tickets in each fare class, starting from the lowest and proceeding in ascending order of price; the demand is assumed to be uniformly distributed throughout each day. 4) PDAwR (Price Daily Assignment with partial Redistribution of unsatisfied demand): a generalization of PDA, in which unsatisfied demand is partially redistributed to the next day. This modeling approach allows us to consider passengers who do not immediately choose a competitor, but await a lower ticket price.

Cite: Brazhnikov E. , Panin A. , Erzin A.
New Dynamic Airline Ticket Pricing Models Based on Combinatorial Optimization
XVI International Conference Optimization and Applications 22-26 Sep 2025