Delay assignment optimization strategies at pre-tactical and - PowerPoint PPT Presentation

Delay assignment optimization strategies at pre-tactical and tactical levels A. Montlaur and L. Delgado Dr Luis Delgado Senior Research Fellow University of Westminster _______________________ Universitat Politècnica de Catalunya Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Overview • • Background Scenario • • Optimization models Results – System overview – Tactical – Stakeholders – Pre-tactical – General ground holding • Conclusions and problem formulation further work – Cost functions Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Background Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Background Tactical traffic management Pre-tactical traffic management Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Background Tactical traffic management Extended region E-AMAN Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Background • RBS Slots Demand • Minimising available passenger F1 S1 delay F2 S2 • Minimising F3 . . S3 . delay . . … . considering … . . . turn-around Fm Sn Optimisation assignment 1 Optimisation assignment 2 Optimisation assignment j Metric 1 V 11 V 12 V 1j Metric 2 V 21 V 22 V 2j Metric i V i1 V i2 V ij Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Optimization models - System overview Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

System overview • Slot window • Slot window between 1 to 3 between 10 to 15 minutes • Optimisation phases minutes • Re-optimised every • Delay realised time a flight enters on-ground at origin the outer radius • Maximum 35 minutes of delay Tactical assigned Tactical uncertainty optimisation E-AMAN Pre-tactical optimisation Origin 500 km Destination 50 km Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Optimization models - Stakeholders Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Stakeholders • Airlines – Flight centric metrics • Passengers – Passenger centric metrics • Importance of optimization function focus Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Optimization models - General ground holding problem formulation Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

General GHP formulation • Deterministic Ground Holding Problem (GHP) • Constraints applied at destination – outer or inner radius M. Ball, C. Barnhart, G. Nemhauser and A. Odoni, Air Transportation: Irregular Operations and Control , Handbook in OR & MS, Vol. 14, 2007. Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

General GHP formulation • Set of intervals • Set of flights • Inputs defined – Capacity at each time interval – Scheduled time of arrival for each flight • Decision variables – If a flight is assigned to arrive at a given time interval (starting at the earliest possible arrival time for that flight) • Problem formulation – Assign flights to intervals minimizing cost (all flights must be assigned, capacity not overpassed) M. Ball, C. Barnhart, G. Nemhauser and A. Odoni, Air Transportation: Irregular Operations and Control , Handbook in OR & MS, Vol. 14, 2007. Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Optimization models - Cost functions Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Cost functions Demand Slots available Four costs models considered ( � �� ) • . . . . . . . . . – GHP Flight: Delay per flight minimised Arrival Scheduled – �� time arrival time Arrival delay Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Cost functions Four costs models considered ( � �� ) • – GHP PAX: Delay per passenger minimised Number of Arrival x passengers �� delay arriving Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Cost functions Four costs models considered ( � �� ) • – GHP Reac: Delay per flight considering reactionary departure delay Subsequent Arrival + 1.8 x departure �� delay delay Latest arrival time – not generate Arrival time departure delay Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Cost functions Four costs models considered ( � �� ) • – GHP Reac Pax: Delay per passengers considering reactionary departure delay Passenger Subsequent + 1.8 x �� arrival passenger departing delay delayed Number of Number of Subsequent Arrival x passengers x passengers departure delay arriving departure delay Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Cost functions Four costs models considered ( � �� ) • – GHP Flight: Delay per flight minimised – GHP PAX: Delay per passenger minimised – GHP Reac: Delay per flight considering reactionary departure delay – GHP Reac Pax: Delay per passengers considering reactionary departure delay Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Scenario and model uncertainty Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Scenario definition and uncertainty Model Sub-model Description Times • Based on 12SEP14 at CDG Scenario Once • Between 5:00 and 11:00 GMT Flight demand • Cancelled flight considered pre-tactically but not tactically • Flights within inner radius excluded Demand Data Repository 2 (DDR2) Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Scenario definition and uncertainty Model Sub-model Description Times • Based on 12SEP14 at CDG Scenario • Between 5:00 and 11:00 GMT Flight demand Once • Cancelled flight considered pre-tactically but not tactically • Flights within inner radius excluded Turnaround Top 10 AC types A320 turn around times Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Scenario definition and uncertainty Model Sub-model Description Times • Based on 12SEP14 at CDG • Between 5:00 and 11:00 GMT Flight demand • Cancelled flight considered pre-tactically but not tactically Scenario • Flights within inner radius excluded Once • AC type for minimum turnaround time (MTT) • AC types top 10 used • Turnaround AC categories otherwise • Burr and Weibull distribution fitting MTT( � )= Max(rand(0.1,0.4),STT( � )) • Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster

Scenario definition and uncertainty Model Sub-model Description Times • Based on 12SEP14 at CDG • Between 5:00 and 11:00 GMT Flight demand • Cancelled flight considered pre-tactically but not tactically • Flights within inner radius excluded Scenario • AC type for minimum turnaround time (MTT) Once • AC types top 10 used • Turnaround AC categories otherwise • Burr and Weibull distribution fitting MTT( � )= Max(rand(0.1,0.4),STT( � )) • Passenger • Triangular distribution between 60-95% centered at 85% demand Fifth SESAR Innovation Days Universitat Politècnica de Catalunya Università di Bologna, Italy, 1 – 3 DEC 2015 University of Westminster