CN-122020387-A - Dual-mode shield machine recognition method based on data driving

CN122020387ACN 122020387 ACN122020387 ACN 122020387ACN-122020387-A

Abstract

The invention discloses a data-driven dual-mode shield card machine identification method, which comprises the steps of collecting tunneling parameters of a plurality of normal tunneling sections and card machine tunneling sections of dual-mode shield engineering, preprocessing data aiming at collected original data to remove invalid data, optimizing tunneling parameters capable of effectively describing the tunneling state of the dual-mode shield machine based on a mathematical statistics method, calculating weight of each tunneling parameter based on an entropy weight method, calculating and obtaining quantitative comprehensive indexes based on weight and cumulative probability density functions of each tunneling parameter, calculating and obtaining a threshold value of a quantitative comprehensive index judgment card machine through the mathematical statistics method, and further obtaining an identification mechanism capable of identifying the tunneling state of the dual-mode shield card machine. The invention can better describe the dual-mode shield tunneling state and can better evaluate and divide the blocking machine tunneling state.

Inventors

WANG QUANWEI
YAN CHANGBIN
GUO JING
ZHAO WEIRONG
LIU JIANLEI
GAO PING
ZHANG TAO
WANG FUQIANG

Assignees

黄河勘测规划设计研究院有限公司

Dates

Publication Date: 20260512
Application Date: 20260204

Claims (8)

1. The data-driven dual-mode shield machine identification method is characterized by comprising the following steps of: s1, collecting original tunneling parameters of a plurality of dual-mode shield engineering normal tunneling sections and a blocking machine tunneling section; s2, preprocessing original tunneling parameters based on a binary discriminant function method and a2σ principle method to remove invalid data; s3, optimizing effective tunneling parameters for effectively describing the tunneling state of the dual-mode shield tunneling machine from the preprocessed tunneling parameters based on a mathematical statistics method; S4, calculating the weight of each effective tunneling parameter based on an entropy weight method; S5, calculating a quantitative comprehensive index based on the weight of each effective tunneling parameter and the cumulative probability density function; S6, calculating a quantitative comprehensive index judgment threshold value of the card machine, and counting a multi-engineering card machine case verification judgment threshold value through a mathematical system; S7, calculating quantitative comprehensive indexes of the tunneling parameters to be evaluated, and identifying the tunneling state of the dual-mode shield card machine according to the judging threshold.
2. The data-driven dual-mode shield machine identification method of claim 1, wherein the tunneling parameters comprise thrust force F, penetration degree p, cutter torque T and cutter rotating speed n, and the binary discriminant function formula is: , , when s=0, the invalid data is judged to be invalid, and the invalid data is removed.
3. The data-driven dual-mode shield machine identification method of claim 1, wherein the 2σ principle is used for eliminating tunneling data except [ mu-2σ, mu+2σ ] according to normal distribution, wherein mu is a mean value and sigma is a standard deviation.
4. The method of identifying dual-mode shield machine based on data driving of claim 1, wherein S3 comprises classifying the preprocessed tunneling parameters according to the normal tunneling section and the machine-blocking tunneling section, and calculating the coincidence of the tunneling parameters x of the normal tunneling section and the machine-blocking tunneling section The formula is: wherein x is a tunneling parameter; 、 the average value of the tunneling parameters x of the normal tunneling section and the average value of the tunneling parameters x of the blocking machine tunneling section are respectively obtained; 、 the variance of the tunneling parameters x of the normal tunneling section and the variance of the tunneling parameters x of the blocking machine are used for determining the degree of coincidence And when the tunneling parameter x is an effective tunneling parameter.
5. The data-driven dual-mode shield machine identification method of claim 1 is characterized in that S4, an evaluation matrix of actual measurement values of each effective tunneling parameter is constructed based on an entropy weight method, entropy values and difference coefficients of each effective tunneling parameter are calculated, and then weights of each effective tunneling parameter are obtained.
6. The method for identifying the dual-mode shield machine based on data driving of claim 1, wherein the quantitative comprehensive index calculation formula is as follows: , wherein, Is a quantitative comprehensive index; The cumulative probability density function of the effective tunneling parameter x of the machine blocking section is provided; The weight of the i-th effective tunneling parameter x.
7. The method for identifying a dual-mode shield machine based on data driving of claim 1, wherein the determination threshold is In the formula (I), in the formula (II), To quantify the comprehensive index Is the overlap ratio of (2); 、 respectively, quantitative comprehensive indexes Average value of normal tunneling section and blocking machine tunneling section; 、 to quantify the comprehensive index And (5) a normal tunneling section and a variance value of the blocking machine tunneling section.
8. The method for identifying a dual-mode shield machine based on data driving of claim 1, wherein the determination threshold is To quantify the comprehensive index And the differential value of the normal tunneling section and the blocking tunneling section.

Description

Dual-mode shield machine recognition method based on data driving Technical Field The invention relates to the technical field of construction evaluation of dual-mode shield tunnels, in particular to a data-driven dual-mode shield machine identification method. Background The tunneling process of the shield machine is essentially a dynamic process of the interaction of the rock mass and the machinery, and tunneling parameters serve as key feedback indexes of the process, are closely related to physical and mechanical properties, stratum structures and the like of surrounding rock mass, and can reflect the working state of tunneling equipment in real time. Currently, tunneling parameters have been widely used for identification of poor geology. However, it is more critical how to realize active early warning and prevention and control of construction disasters based on tunneling parameters. Because construction disasters are frequently generated in poor geology or high-risk stratum, the response rule of tunneling parameters under the geological conditions is deeply analyzed, indicative data characteristics are extracted, and then a disaster early warning mechanism based on the tunneling parameters is constructed, so that the method has important significance for guaranteeing construction safety. In the shield method construction, common disasters comprise earth surface subsidence, collapse, gushing water, surrounding rock large deformation, machine blockage and the like. The machine blocking disaster has special positions, namely the machine blocking disaster is not only the final result possibly caused by disasters such as collapse, large deformation of surrounding rock and the like, but also the direct disaster is reflected after the interaction of the rock mass and the shield machine is unbalanced. Therefore, from the view point of a typical disaster of a machine blocking, the system analyzes evolution rules of tunneling parameters before, during and after occurrence, deep extracts characteristic parameters capable of representing risks of the machine blocking, builds a corresponding risk identification and early warning model, has great value in improving safety, efficiency and intelligence level of shield construction, and is a core technical direction of urgent research at present. Disclosure of Invention The invention aims to provide a data-driven dual-mode shield machine identifying method, which is used for solving the problem of delayed machine blocking disaster early warning and dependence on manual experience in the existing shield construction based on tunneling parameter real-time data and constructing a machine blocking risk early identifying method. In order to achieve the above purpose, the data-driven dual-mode shield card machine identification method based on the invention comprises the following steps: s1, collecting original tunneling parameters of a plurality of dual-mode shield engineering normal tunneling sections and a blocking machine tunneling section; s2, preprocessing original tunneling parameters based on a binary discriminant function method and a2σ principle method to remove invalid data; s3, optimizing effective tunneling parameters for effectively describing the tunneling state of the dual-mode shield tunneling machine from the preprocessed tunneling parameters based on a mathematical statistics method; S4, calculating the weight of each effective tunneling parameter based on an entropy weight method; S5, calculating a quantitative comprehensive index based on the weight of each effective tunneling parameter and the cumulative probability density function; S6, calculating a quantitative comprehensive index judgment threshold value of the card machine, and counting a multi-engineering card machine case verification judgment threshold value through a mathematical system; S7, calculating quantitative comprehensive indexes of the tunneling parameters to be evaluated, and identifying the tunneling state of the dual-mode shield card machine according to the judging threshold. Further, the tunneling parameters comprise thrust force F, penetration degree p, cutter torque T and cutter rotating speed n, and the binary discriminant function formula is as follows:,, when s=0, the invalid data is judged to be invalid, and the invalid data is removed. Further, the 2σ principle is used for eliminating tunneling data except [ mu-2σ, mu+2σ ] according to normal distribution, wherein mu is the mean value, and sigma is the standard deviation. Further, the step S3 comprises the specific steps of classifying the preprocessed tunneling parameters according to the normal tunneling section and the blocking tunneling section, and calculating the coincidence ratio of the tunneling parameters x of the normal tunneling section and the blocking tunneling sectionThe formula is: wherein x is a tunneling parameter; 、 the average value of the tunneling parameters x of the normal tunneling section and the a