A Multi-Objective Optimization Framework for Space Transportation Systems Based on Mixed-Integer Linear Programming and Monte Carlo Simulation

Abstract

This paper establishes a multi-objective optimization analysis framework for large-scale transportation system planning. Based on a multi-stage mixed-integer linear programming (MILP) model, it models the system under constraints of transport capacity, infrastructure scale, and operational conditions, with transportation costs and project duration as primary optimization objectives. By introducing the ε-constraint method, the multi-objective problem is transformed into a single-objective solution, and a Pareto frontier between cost and construction duration is constructed to identify optimal configuration schemes under different transportation strategies. To extend the model, operational factors such as maintenance cycles, system capacity ramp-up, and learning curves are incorporated to dynamically characterize changes in transportation capacity and costs. A digital twin analysis framework is further developed by integrating discrete-event simulation with Monte Carlo simulation to evaluate random disturbances like equipment failures and transportation delays. The Conditional Value at Risk (CVaR) method is employed to measure system risks under extreme scenarios. Concurrently, rolling time-domain optimization enables dynamic adjustments to transportation decisions. Research findings demonstrate that this methodology achieves coordinated optimization and risk control for transportation systems under complex constraints and uncertainty, providing a universally applicable optimization modeling and simulation analysis approach for planning complex engineering transportation systems.