摘 要

近年来，盾构施工技术正扮演着越来越重要的作用，已经逐渐取代矿山法，成为地铁隧道建设的首选工法，随着我国各大中小城市纷纷开始修建地铁线路，盾构的使用范围也更广。盾构作为一种机械化程度相当高的工法，仍然无法避免目前我国建筑业施工粗放，效率不高等特点，一台盾构每天的能耗量相当于一座小型工厂，其中刀盘消耗的能耗占了全部的一半以上。盾构刀盘能耗符合当前各行各业节能减排的趋势，另一方面也可以有效的提高盾构机工作效率，减少工期，节约工程施工成本，确保施工质量和安全。

由于盾构在掘进过程中，除了刀盘驱动系统外的其他系统工作时的功率相对比较稳定，而刀盘驱动系统要考虑到盾构刀盘的脱困问题，因此在设计刀盘驱动系统时的额定功率往往很高，但当盾构机进入正常掘进的状态时，此时的刀盘驱动功率远小于额定功率，因此长期处于欠负载的状态工作，因此效率不高，有着巨大的节能潜力。盾构机在工作时，工作面的水文地质条件的变化会直接影响盾构刀盘的能耗，但由于直接实时监测工作面的地质变化比较困难，因此采用通过盾构机的掘进参数的来预测刀盘能耗。盾构机的各项掘进参数是盾构当前工作状态下的直接表现，工作面地质条件的变化都会引起相应的掘进参数发生变化，因此通过盾构机掘进参数来预测刀盘能耗是可行的。

为此，本文总结最近的有关盾构掘进的相关文献和以往的施工经验，确定了推力、刀盘转速、扭矩、贯入度、掘进速度、土压力和气仓压力这七个参数作为建立盾构刀盘能耗预测模型的输入参数。为了提高模型的预测精度和确保模型的稳定性，本文采用最小支持向量机的算法建模，同时采用BP神经网络方法对同样的数据进行预测，根据结果的准确性来说明该方法的适用性，最后的预测结果发现利用SVM算法建模误差更小而且耗时更少，说明该方法可行。不过本文还存在一些不足之处，例如对SVM建模过程中参数的选择和掘进参数对能耗的影响都还可以进一步研究。

关键词：土压平衡盾构；掘进参数；支持向量机；能耗

Abstract

In recent years, shield construction technology is playing an increasingly important role, has gradually replaced the mine law, a subway tunnel construction of the preferred method, with China's major cities have started construction of subway lines, shield use Wider Shield as a very high degree of mechanization of the law, still can not avoid the current construction industry in China extensive, low efficiency and so on, a shield energy consumption per day is equivalent to a small factory, which consumes the energy consumption of the cutter More than half of all. Shield cutter energy consumption in line with the current trend of energy-saving emission reduction in all walks of life, on the other hand can also effectively improve the efficiency of the shield machine, shorten the construction period, saving construction costs, to ensure the quality and safety of construction.

As the shield in the tunneling process, in addition to the cutter drive system other than the system when the power is generally relatively stable, and the cutter drive system to take into account the shield cutter out of the problem, so the design of the cutter drive system The rated power is often high, but when the shield machine into the normal state of the tunnel, this time the cutter drive power is much smaller than the rated power, so long-term under-load state of work, so the efficiency is not high, has a huge energy-saving potential. When the shield machine is working, the change of the hydrogeological conditions of the working face will directly affect the energy consumption of the shield cutter. However, it is difficult to monitor the geological changes of the working face directly. Therefore, this paper adopts the parameters of the tunneling To predict the energy consumption of the cutterhead. The excavation parameters of the shield machine are the direct performance under the current working condition of the shield, and the change of the geological conditions of the working face will cause the corresponding tunneling parameters to change. Therefore, it is feasible to use the shield tunneling parameters to predict the energy consumption of the cutterhead.

For this reason, this paper summarizes the relevant literatures on shield tunneling and the previous construction experience to determine the thrust, cutter speed, torque, penetration, tunneling speed, earth pressure and gas pressure. The input parameters of the toolhead energy consumption prediction model. In order to improve the prediction accuracy of the model and ensure the stability of the model, this paper uses the minimum support vector machine to model the algorithm, and uses the BP neural network method to predict the same data. According to the accuracy of the results, The results of the final prediction show that the SVM algorithm is smaller and less time-consuming, which indicates that the method is feasible. However, there are some shortcomings in this paper, such as SVM modeling process parameters and the choice of the parameters of the impact of energy consumption can also be further studied.

Key word:Earth pressure balance shield；Tunneling parameters；Support vector machine；energy consumption