An Optimization Framework for Multimodal Transportation Based on Mixed-Integer Nonlinear Programming and Monte Carlo Simulation

Abstract

This paper proposes a multimodal transportation optimization framework for large-scale Earth–Moon cargo delivery. The problem addresses resource allocation, uncertainty, and multi-constraint coupling in high-dimensional decision spaces. To tackle these challenges, a unified optimization framework combining mixed-integer nonlinear programming (MINLP) and stochastic simulation is developed. The model integrates transportation capacity modeling, cost evaluation, and decision optimization within a single computational pipeline. Uncertainty is characterized through probabilistic modeling, while Monte Carlo simulation is employed to evaluate system performance under stochastic disturbances. Experimental results demonstrate that the proposed hybrid strategy significantly outperforms single-mode approaches. The system completion time is reduced from approximately 4200 years to 1726 years, with total cost reduced to 4.87×1011 USD. Under uncertainty, the probability of completing tasks within 2000 years exceeds 60%, indicating improved robustness. The proposed framework provides a general and scalable methodology for complex transportation and resource allocation problems under uncertainty.