复合地层大直径盾构刀具磨损预测与换刀决策研究

张婉

摘  要

随着城市地下空间开发的大规模展开，隧道工程建设项目不断增加，盾构工法以其具有机械化作业程度高、安全可靠的优点而被广泛采用。近年来大直径盾构隧道项目在全国范围内快速增长，导致工程面临的地质情况越来越复杂，特别是软硬相间复合地层工况增多，施工风险以及换刀频率明显增加。由于盾构刀具与岩土体持续作用产生磨损，面对复杂的地质条件，更加容易造成工程掘进效率降低、意外停机进而导致综合成本陡增情形的出现。因此，建立准确的刀具磨损预测模型，以选择合适的开仓换刀时机，成为大直径盾构施工中亟待解决的一个工程难题。

为了实现盾构的安全高效掘进，本文基于团队自主研发的盾构工程数字化管控系统积累的大量工程案例的有关掘进参数和地质条件参数，并考虑相关参数间的相关性，构建了刀具磨损影响因素的输入-输出预测模型，实现了复合地层单刀单环磨损量预测。基于磨损研究对刀具寿命进行预测，提出基于项目进度以及成本的不同维修策略。主要研究工作包括：

（1）研究了复合地层条件刀具磨损影响因素及其相关关系。得出了磨损预测中最重要的影响因素即地质参数，为刀具磨损预测模型的指标选取提供理论依据。同时对盾构机刀具配置进行了统计分析，计算了不同地层条件下刀具的磨损系数。

（2）引入XGBoost机器学习算法，建立了复合地层刀具磨损预测模型。模型以掘进参数、结构参数和地质参数作为输入，预测刀具单刀单环的磨损量作为输出。采用该输入-输出指标体系模型的R2精度能达到0.90，表明该模型预测效果较好。

（3）定义了刀盘磨损度指标，构建了刀盘磨损度与盾构机掘进效率的数学模型。结合刀具磨损系数分析进行刀具寿命预测，为预防性维修策略的制定提供依据，利用AnyLogic软件模拟刀具磨损和盾构施工过程，从项目进度和成本两个角度对不同换刀方案进行评估，得到最小成本的停机方案。

本文研究为大直径盾构在复合地层中掘进的刀具磨损预测与换刀决策，提供了量化分析依据，算法成果已在某市地铁区间隧道盾构工程中得以试点应用。

关键词：盾构隧道；复合地层；刀具磨损预测；换刀决策；维修策略

Abstract

With the large-scale development of urban underground space and the continuous increase of tunnel construction projects, the shield tunneling method has been widely used for its advantages of higher degree of mechanization, safety and reliability. In recent years, the large size shield tunnel project has been expanded nationwide, and the shield has encountered more complex geological conditions, such as soft-hard composite stratum, resulting in a significant increase in construction risks and frequent cutter change. The cutting tools of shield tunnel machine endure substantial wear in excavation, where the geological conditions are complicated and uncertain. It can lead to a reduction in excavation efficiency, disruption of construction schedules, and extra costs. Therefore, it is important to establish an accurate prediction model of cutter wear, solving the problem of selecting the appropriate opening time for large size shield tunnel.

In this paper, the prediction model of cutter wear was proposed, based on the relevant tunneling parameters and geological parameters provided by the digital management and control system. And the correlation of the relevant parameters was considered, as to predict the cutter wear per ring in composite stratum. Cutter life was calculated by previous cutter wear prediction model, and different maintenance strategies based on project schedule and cost assessment were proposed. The main research work is as follows:

(1) The correlation of cutter wear influencing factors in composite stratum was studied. It is pointed out that the most important influencing factors in cutter wear prediction are geological parameters, which provided theoretical basis for the selection of indicators of cutter wear prediction model. In addition, the configuration of the cutter head was statistically analyzed, and the wear coefficients of cutting tools under different geological conditions were calculated.

(2) The XGBoost algorithm was used to establish the cutter wear prediction model in composite stratum. The tunneling parameters, structural parameters and geological parameters were taken as input variables to predict the wear of single cutting tool per ring. The accuracy of the input-output index system model can reach 0.90, indicating that the algorithm has good applicability in wear prediction.

(3) The index of cutterhead wear degree was defined, and the mathematical model between cutterhead wear degree and shield tunneling efficiency was constructed. Combined with cutter wear coefficient analysis, cutter life prediction was carried out, which provided a basis for the formulation of preventive maintenance strategies. The AnyLogic software was used to simulate the process of cutter wear and shield construction, and different tool maintenance strategies were evaluated from the perspective of project progress and cost.

This study provided a quantitative analysis basis for cutter wear prediction and cutter change decision of large size shield tunnel in composite stratum, and the results of the algorithm have been applied in a city subway tunnel shield project.

Key words: Shield tunneling; Composite stratum; Cutter wear prediction; Cutter change decision; Maintenance strategy