Research on Airline Crew Scheduling Model for Fatigue Management

Crew fatigue is a major cause of aviation accidents. Current civil aviation crew scheduling relies on civil aviation regulations, managing fatigue via restricting flight, duty and rest time; however, fatigue is also affected by time difference, circadian rhythm and mental fatigue severity in different work modes, which makes time-only constraints insufficient for scientific fatigue measurement and management and disregards safety hazards. Thus, integrating fatigue management into crew scheduling has practical value. This paper studies fatigue-oriented crew scheduling by first establishing a GA-LSSVM-based multiple regression model to predict crew fatigue in different flight loop tasks—achieved by exploring the quantitative relationship between pilot fatigue and key influencing factors—then investigating fatigue-considered crew ring formation, building a multi-objective optimization model and verifying its effectiveness via the NSGA-II algorithm, and finally studying fatigue balance-oriented task assignment, constructing a multi-objective model, solving the optimal scheme with the NSGA-II algorithm. Results show the proposed method reduces fatigue by 2.86% (based on the paper’s subjective evaluation data) in airline operations, with only a 0.28% increase in operating costs, while also improving crew satisfaction and ensuring fatigue balance.