CN-120335002-B - Slag soil intelligent treatment method and system for advanced geological prediction of combined tunnel

CN120335002BCN 120335002 BCN120335002 BCN 120335002BCN-120335002-B

Abstract

The invention discloses an intelligent muck treatment method for advanced geological prediction of a combined tunnel, which comprises the steps of obtaining a seismic wave training set, carrying out geological analysis on muck types in the seismic wave training set, obtaining dynamic calculation parameters and physical data corresponding to each muck type, corresponding to the seismic wave characteristics and the dynamic calculation parameters and the physical data, obtaining an updated seismic wave training set, training a neural network by using the seismic wave training set to obtain a muck prediction treatment model, obtaining real-time geological prediction data, preprocessing the real-time geological prediction data, inputting the preprocessed real-time geological prediction data into the muck prediction treatment model, outputting the muck type through the muck prediction treatment model, and generating muck treatment suggestions by combining the working state and geological prediction of a shield machine so as to treat muck generated in the shield construction process. The method can predict the type of the slag soil generated during tunneling in time, and improves the timeliness of prediction.

Inventors

BAO DEYONG
LI DA
LIN CHAOFENG
LIU TENG
HE QIHAI
LUO YONG
SUN WEIXIN
ZHANG WENGANG
YANG YANG

Assignees

中铁第四勘察设计院集团有限公司
厦深铁路广东有限公司
中铁二局集团有限公司
重庆大学
重庆大学溧阳智慧城市研究院

Dates

Publication Date: 20260512
Application Date: 20250317

Claims (8)

1. An intelligent slag soil treatment method combined with advanced geological prediction of a tunnel is characterized by comprising the following steps: s1, acquiring a seismic wave training set; The seismic wave training set comprises seismic wave characteristics and slag soil categories corresponding to each seismic wave characteristic; s2, performing geological analysis on the slag soil categories in the seismic wave training set, obtaining dynamic calculation parameters and physical data corresponding to each slag soil category, and corresponding the seismic wave characteristics to the dynamic calculation parameters and the physical data to obtain an updated seismic wave training set; S3, training the neural network by using the seismic wave training set to obtain a muck prediction processing model; S4, acquiring real-time geological forecast data, preprocessing the real-time geological forecast data, inputting the preprocessed real-time geological forecast data into the muck prediction processing model, and outputting muck types through the muck prediction processing model; S5, generating a muck treatment suggestion according to the muck type and combining the working state and geological forecast of the shield machine so as to treat muck generated in the shield construction process.
2. The method of claim 1, wherein after S1, the method further comprises: preprocessing the seismic wave training set to obtain a preprocessed seismic wave training set; The step S2 specifically comprises the following steps: And carrying out geological analysis on the slag soil category in the preprocessed seismic wave training set, obtaining dynamic calculation parameters and physical data corresponding to each slag soil category, and corresponding the seismic wave characteristics to the dynamic calculation parameters and the physical data to obtain an updated seismic wave training set.
3. The method of claim 2, wherein in S2, the kinetic calculation parameters include longitudinal wave velocity, transverse wave velocity, wave velocity ratio, density, poisson 'S ratio, young' S modulus, and bulk modulus; The physical data includes porosity, moisture content, and particle size distribution.
4. The method of claim 3, wherein in S3, the muck prediction processing model comprises an input layer, a convolution layer, a flattening layer, a full connection layer, and an output layer; The data form acquired by the input layer is N8*3; the convolution layers comprise a first convolution layer, a first pooling layer, a second convolution layer, a second pooling layer, a third convolution layer, a third pooling layer, a fourth convolution layer and a fourth pooling layer; The number of convolution kernels of the first convolution layer is 64, the number of convolution kernels of the second convolution layer is 128, the number of convolution kernels of the third convolution layer is 256, the number of convolution kernels of the fourth convolution layer is 512, the convolution kernels of the first convolution layer, the second convolution layer, the third convolution layer and the fourth convolution layer are 3*3, and the activation functions of the first convolution layer, the second convolution layer, the third convolution layer and the fourth convolution layer are ReLU functions; the pooling window sizes of the first pooling layer, the second pooling layer, the third pooling layer and the fourth pooling layer are all 2 x 2; the full-connection layer comprises a first full-connection layer and a second full-connection layer; The number of nodes of the first full-connection layer is 1024, and the number of nodes of the second full-connection layer is 512; the activation functions of the first full-connection layer and the second full-connection layer are ReLU functions; the Dropout rates of the first full-connection layer and the second full-connection layer are 0.5; the number of the nodes of the output layer is consistent with the number of the slag soil categories.
5. The method according to claim 4, wherein in S3, the loss function of the muck prediction processing model is Categorical Crossentropy; the optimizer of the muck prediction processing model is an Adam optimizer, and the initial learning rate of the Adam optimizer is 0.001; L2 regularization is applied in the fully connected layer, and the weight attenuation coefficient is 0.001.
6. An intelligent slag processing system for advanced geological prediction of a combined tunnel, which is realized based on the intelligent slag processing method for advanced geological prediction of the combined tunnel according to any one of claims 1-5, and is characterized by comprising the following steps: The model training module is used for obtaining a seismic wave training set, carrying out geological analysis on the slag soil category in the seismic wave training set, obtaining dynamic calculation parameters and physical data corresponding to each slag soil category, and corresponding the seismic wave characteristics with the dynamic calculation parameters and the physical data to obtain an updated seismic wave training set; the seismic wave training set comprises seismic wave characteristics and slag soil categories corresponding to each seismic wave characteristic; The real-time data acquisition module is used for acquiring real-time geological forecast data, preprocessing the real-time geological forecast data, inputting the preprocessed real-time geological forecast data into the muck prediction processing model, and obtaining the muck type generated in the shield construction process; The muck analysis module is used for acquiring real-time geological forecast data, inputting the real-time geological forecast data into the muck prediction processing model and obtaining the muck type generated in the shield construction process; and the muck treatment module is used for generating muck treatment suggestions according to the muck type and by combining the working state and the geological forecast of the shield machine so as to treat muck generated in the shield construction process.
7. The intelligent slag processing system of claim 6, wherein said real-time data acquisition module comprises a seismic wave exciter and a detector; the seismic wave exciter is used for exciting active source seismic waves, and the detector is used for receiving all-space vibration signals.
8. The intelligent slag treatment system of claim 7, the slag soil intelligent treatment system is characterized by further comprising: and the data preprocessing module is used for preprocessing the seismic wave training set to obtain a preprocessed seismic wave training set.

Description

Slag soil intelligent treatment method and system for advanced geological prediction of combined tunnel Technical Field The invention relates to the technical field of shield tunnel excavation, in particular to an intelligent slag soil treatment method and system for advanced geological prediction of a combined tunnel. Background Along with the gradual deepening of the urban process in China, the development and utilization of the urban underground space become an important direction of urban development. Compared with the traditional tunnel excavation method, the shield machine integrates various operations such as excavation, support, propulsion, stacking and the like, has obvious construction efficiency advantages and better cost benefits, does not need to be removed on the ground in a large area in the construction process of the shield machine, does not block traffic, has no noise in construction, does not subside on the ground, does not influence the normal life of residents, and has obvious environmental protection and construction safety advantages. Therefore, the technology is considered to be a common technology in the field and is widely applied to the fields of major engineering and infrastructure such as urban rail transit, underground pipe gallery, railway and highway tunnel and the like. Tunneling a tunnel using shield technology tends to produce a large amount of shield slag. At present, in domestic tunnel engineering, the treatment of shield slag soil is still mainly carried out by a traditional stacking landfill method, the occupied area is large, and the appearance of cities is influenced. In the shield tunnel construction process, in order to protect a cutter head of a shield machine, reduce the torque of the cutter head and increase the stratum flow plasticity, a surfactant, commonly known as a foaming agent, is added into a soil bin. The foaming agents can be adsorbed in the dregs, so that the water content of the foaming agents is high, the foaming agents are not easy to transport, a series of physical and chemical reactions also occur, and the difficulty of environment-friendly treatment of the dregs is increased. If the water is directly discharged without proper treatment, soil, surface water and underground water are polluted, and the ecological environment is greatly damaged. And the difference of stratum properties of the excavation surface can lead the shield to generate various different types of dregs with different parameters of grain composition, penetration index, water content, flowing state and the like, so that the required treatment schemes are also different. The intelligent construction of tunnels is one of research hotspots in the current railway tunnel construction field. The shield method is industrial operation, the construction process is organized according to the assembly line, the degree of mechanization is high, and the method has the foundation and the advantages of realizing the intellectualization. However, the current intelligent technology is mainly focused on the visual, digital and intelligent centralized management in aspects of tunnel investigation, design and construction, and the whole process intelligent technology of tunnel construction combining investigation and construction waste treatment is rarely available. In modern tunnel engineering, the technology for identifying the muck has important significance for optimizing the construction process, improving the operation efficiency and guaranteeing the construction safety, and at present, the muck identification mainly depends on equipment such as computer vision, a cutterhead sensor, a rheometer and the like. The computer vision technology can realize preliminary judgment of the geological conditions in front through real-time analysis of the face image, and the cutterhead sensor can monitor various parameter changes under the working state of the cutterhead so as to indirectly reflect the geological characteristics of the face. As for the application of rheometers, the measurement of the physical properties of the excavated muck, such as viscosity, density, etc., is limited primarily, and this information helps to understand the state of muck in the earth bin, but does not provide direct data on the unexcavated area. However, the above methods have a common limitation in that they focus mainly on the analysis of the current or completed work area, lacking the ability to foresight prediction of the forward geological situation. This lack of timeliness limits the effectiveness of the muck identification technique in guiding muck treatment protocols. Especially under complex geological conditions, timely and accurate prediction of the forward geological conditions is critical for preventing potential risks. In recent years, with the development of artificial intelligence technology, advanced geological prediction based on machine learning has become possible. According to the technology, multipl