Intelligent optimization framework for tunnel construction organization driven by multi-source data and process-level simulation

Tunnel construction faces challenges including complex geological conditions, strong inter-coupling among processes, and frequent dynamic disturbances. Existing organization methods struggle to achieve dynamic multi-objective optimization of quality, schedule, cost, and safety. This paper proposes an intelligent optimization framework integrating multi-source information with process-level simulation, which decouples spatiotemporal coupling in complex systems through three-level hierarchical modeling. By combining disturbance prediction with simulation calibration, it enables dynamic adjustment of plans and multi-objective optimization. First, a parametric process model is built using hierarchical modeling, with dynamic algorithms simulating random disturbances. Second, expert evaluation and simulation calibration establish mapping relationships between quality scores and production data. Finally, a closed-loop system of "planning–execution–monitoring–optimization" is formed through dynamic feedback. This research contributes: (1) A spatiotemporal-decoupled hierarchical modeling method overcoming coupling limitations of static models; (2) Collaborative simulation for disturbance events enhancing prediction accuracy through joint modeling of cycles and maintenance strategies; (3) A dual-driven mechanism integrating expert knowledge with simulation data, enabling adaptive adjustment of multi-objective weights.