A linear programming framework for optimizing molecular combinations of asphalt four components
摘要
To enhance the efficiency and accuracy of asphalt molecular modeling, this investigation proposes a four-component molecular combination optimization method for asphalt based on linear programming. This method automatically selects the optimal representative molecular combination while satisfying multiple constraints, such as component ratios, element mass fractions, and hydrogen-to-carbon atomic ratio. Additionally, this investigation has developed a supporting graphical modeling system based on the PuLP library and Streamlit framework, featuring visual operations, flexible parameter configuration, and result export. Case studies demonstrate that the model output is highly consistent with experimental values, with all index errors within ± 2%. The system demonstrates well-fitting ability, solution stability, and engineering practicality, providing an efficient and intelligent solution for asphalt molecular simulation and formulation optimization.