详情

原文传递 Taxi Planning Optimization using Mixed Integer Programming Memorandum M-799

题名：	Taxi Planning Optimization using Mixed Integer Programming Memorandum M-799
作者：	M.A.R. van Velthuizen
关键词：	A-SMGCS, taxi planning, linear programming, mixed integer programming
摘要：	This report covers the development of an off-line taxi planning model for the Advanced Surface Movement Guidance and Control System (A-SMGCS). This system is currently under development to enhance the efficiency of the airport ground movements, while maintaining a high level of safety in all weather conditions. The output of the taxi planning model is an output file that can be used as the input file for the Simulator for Airport Surface Traffic (SIMAST). The programming techniques used for the taxi planning model are Linear Programming and Mixed Integer Programming. It is solved using CPLEX which is a tool designed for solving such techniques. The requirements which the model has to meet are given and explained in this report. The model can minimise three values (the total taxi time and the total waiting time of all aircraft and the difference between the actual and planned time of arrival at the takeoff runway for all departing aircraft) and maximize one value (the number of aircraft reaching the takeoff runway on the planned time). Because weighting factors arc used, the values can be minimized or maximized separately or a combination of two or more values. Also the level of importance of each value can be chosen. To optimize the taxi planning, there are three management instruments: rerouting of the aircraft, variation of the taxi speed of the aircraft and waiting of the aircraft at the APRON or at some predetermined areas on the airport. To test the taxi planning model, a prototype airport has been designed. This is a non-existing airport, but it has some characteristics of Amsterdam Airport Schiphol. Because multiple values can be minimized or maximized, four scenarios were written. Each scenario minimizes or maximizes different values. After calculations of these scenarios it can be concluded that Mixed Integer Programming techniques are very useful to solve taxi planning problems. The test point out, that the number of variables restricted to be integer are very small and therefore also the calculation times are very small. The model is also very flexible and can easily be adjusted to minimize or maximize a different (combination of) value(s). Further, the preference for one of the management instrument depends on the weighting factors.
报告类型：	科技报告