To explore the coupling among the safety, economic, and environmental objectives in flight operations control, the trade-off patterns among these objectives were analyzed. The core scope was defined, key factors and coupling feedback relationships were identified, and a system dynamics (SD) model with seven stocks was constructed, based on which a semi-quantitative trade-off analysis was conducted. A fully connected neural network was then introduced to fit the nonlinear mapping between six policy levers and four core objectives, and a multi-objective optimization model was built using the elitist Non-dominated Sorting Genetic Algorithm II (NSGA-II), generating the Pareto front. The constraint relationships were quantified via dynamic trend analysis, threshold attainability tests, and compromise and trade-off coefficients, and the stability of the conclusions was verified by robustness tests. The mean compromise coefficient was 0.957, and the mean profit–safety, carbon emissions–profit, and carbon emissions–safety trade-off coefficients were 12.335, 4.757, and −4.110, respectively, indicating trade-offs among the safety, economic, and environmental objectives. The study finds that under a given total resource budget, policy space, and the airline’s existing operating mode, the three objectives cannot simultaneously attain their individual optima within the feasible solution set. A strong trade-off exists between safety and economy, while the impacts of environmental performance on the other two objectives are constrained by policy settings and technological pathways, offering theoretical references for airlines’ operational decision-making.