Optimization Design Research On Airline Flight Scheduling

Airline flight scheduling is a very hard and important work, because it is the core elements of airlines and the basis of any other operation planning which relates to the flight scheduling and ensures the flight planning's successful implementation. The essence of flight scheduling is to optimize the configuration of primary airline's resources by organizing reasonably and planning precisely. Its quality and efficiency not only are concerned with the safety of operation, but also are the important trump of reducing airline's operation costs and improving competitive power. Based on analyzing the situation of domestic airline opertational planning system, this dissertaton determines the study subject as flight frequency and fleet assignment.This paper includes two main aspects, and the first part is research on flight frequency. In this part, firstly, we propose a combined forecast approach by combining the multiple seasonal ARIMA model with the seasonal grey system model based on analyzing the character of airline passenger flow. Then we put forward a market share model on the object airline in the competitive circumstance. So the air route passenger flow (O-D passenger flow) of the object airline is obtained according to these two models. At last a flight frequency model is founded according to the air route passenger flow and the aircraft supply of the object airline. The second one is study about fleet assignment problem. First, we present weekly fleet assignment model in order to assigning the only fleet to the scheduled flight leg according to the characters of domestic route network and flight scheduling. Based on this, we develop weekly fleet assignment model with homogeneity, the results show the homogeneity rate is at the cost of airline profit. So we can balance the homogeneity rate and airline profit before making decision.The primary innovations are as follows:1. On the basis of analyzing air route passenger flow's non-linear character which combines increase trend with seasonal fluctuation, a combined forecast model forecasting the air route total passenger flow is constructed. The model not only simulates the increase trend of passenger flow, but also does well in the seasonal fluctuation.2. We analyze the factors that affect market share of the object airline in fixed total passenger flow and the relationship among these factors. According to explanative variables such as flight frequency, fleet index and the average fare, we present an object airline market share model in the competitive circumstance. So we can forecast the object airline passenger flow of route by combining the market share model with seasonal combined forecasting model.3. There are some disadvantages in the research on flight frequency in the past scholar's papers, so we propose a flight frequency model whose objective function is profit maximization and restraint condition is supply of airline in the competitive circumstance.4. We develop the time-space network of basic fleet assignment according to the characters of domestic airline network and flight scheduling, and present the notion of false flight arc and false flight, then weekly fleet assignment model is founded. When solving the model, we present the special ordered set branch-and-bound algorithms according to the trait of the model's decision matrix.5. We introduce the notion and the significance of fleet homogeneity, and then we research on the weekly fleet assignment further by bringing forward weekly fleet assignment model with homogeneity.