Time–Cost Trade-Off Optimization in Ultralong Tunnel Construction Based on Working Face Planning

Ultralong tunnels are a critical component of constructing transportation infrastructure in mountainous regions. The construction period of these tunnels has a direct impact on the overall process of opening the line to traffic and operation. Although parallel construction technology with multiple working faces can effectively shorten the construction period, it leads to a significant increase in costs related to labor allocation, equipment deployment, and the construction of ancillary tunnels. This study proposes a mixed-integer quadratic programming (MIQP) model to address the challenge of dynamically optimizing the balance between construction time and costs. The model is based on the cost-differentiated characteristics of different tunneling directions during tunnel construction and integrates decision variables related to construction directions. An adaptive, nondominated sorting genetic algorithm II with an elite retention strategy (ANSGA-II-ERS) is subsequently implemented to address the model. This optimization model effectively addresses the issues of low efficiency and suboptimal economic performance inherent in traditional manual decision-making methods, thereby enriching the theoretical framework for optimizing the construction schedules of ultralong tunnels. The verification results from practical engineering cases demonstrate that compared with the original construction plan, under the constraints of the construction period and resources, the plan optimized by this model successfully reduced one ancillary tunnel and two construction working faces, significantly enhancing the overall project benefits. The findings of this research endeavor can provide project decision makers with a scientific and quantifiable basis for selecting a plan. This study has significant theoretical value and broad engineering application prospects for ultralong tunnel projects with various geological conditions and construction scales.