Fleet Assignment Problem Research Based On Fuzzy Multi-objective Optimization

Aircraft scheduling program is one of the most basic but most important production plan for the airline company, the quality of scheduling is directly related to the operation efficiency and revenue of the airline company in a period of time. With the rising number of Major airlines' fleet in recent years, aircraft scheduling rely solely on traditional handwork becomes more difficult, at the same time, rapid flight scheduling in emergencies is also facing a severe test, so aircraft scheduling automation is the current domestic airline problems to be solved urgently.The author considers the status of the aircraft scheduling of the domestic airline, studies the aircraft scheduling algorithms and models, then applies the fuzzy theory into the scheduling of aircraft, combines fuzzy theory with the diversified targets to be optimized in aircraft scheduling, the writer proposes a fuzzy multi-objective optimization method of aircraft scheduling and build a software platform using the current Air China flight schedule for the actual calculations.Aircraft scheduling program is a famous NP-hard problem of the civil aviation sector, it is a combinatorial optimization problem of multi-objective and multi-constraints. Reasonable fleet assignment not only helps the aircraft operate safety and punctuality, but also improves the utilization of the fleet and effectively reduces the costs of operating and maintenance. The optimization objectives of fleet assignment problem include maximizing profits, balanced using of aircraft and flights fast recovery; Constraints include flight coverage, maintenance requirements and fleet equalization.This paper presents a method of hierarchical model for the aircraft scheduling optimization. Mainly involves two levels of optimization: Firstly, use a model to minimize the using number of aircraft, this can constructs minimum flights strings which could cover all the flights each day, and second, use a model to optimize the properties of flight loop, this could cover all the flights strings and get the weekly path of the airplane. In the second layer optimization, the paper use a single objective optimization model for the first to discuss and simulate the theory and algorithms of the hierarchical model in detail, the experiment results demonstrate that hierarchical optimization method can effectively reduce the use of aircrafts and reduce the number of constraints in the model, the problem scale can be reduced, these makes a number of current fleet assignment become possible. And then the author studies the multi-objective model of scheduling aircraft on the simulation platform. Firstly, calculate theoptimal solution of each single objective in the multi-objective, then use the maximum and minimum model to modify the weight of each objectives and the importance of each objectives, make the optimal solution of the multi-objective model could cover as many optimal solution of single-objective models as possible, this reflects the superiority of fuzzy multi-objective model dynamically select targets to be optimized by using the fuzzy factor.