TBM tunneling parameter abnormal data identification method and system

Abstract

The invention discloses a TBM tunneling parameter abnormal data identification method and system, and belongs to the technical field of data processing. According to the method, a polynomial chaos expansion regression model is constructed through polynomial chaos expansion, and abnormal data in TBM tunneling parameter data are identified by using data clustering and polynomial chaos expansion regression errors; the correlation between TBM tunneling parameters is described through polynomial chaos expansion, and differences between data are compared through the correlation between the TBM tunneling parameters; constructing a clustering objective function based on a polynomial chaos expansion regression error, and performing optimization solution on the clustering objective function by using a Lagrange multiplier method to obtain a TBM tunneling parameter membership matrix; whether the data are abnormal or not is judged through the TBM tunneling parameter membership degree matrix, and accurate identification of the TBM tunneling parameter abnormal data is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing applied to TBM excavation, and in particular relates to a method and system for identifying abnormal data of TBM excavation parameters. Background Art

[0002] TBM (hard rock tunnel boring machine) is a major equipment used in hard rock tunnel construction. It has the advantages of high excavation efficiency, good safety, and good environmental friendliness. It is widely used in tunnel excavation construction. During the tunnel construction process, TBM collects relevant data such as thrust, torque, pressure, temperature, and current through sensors, and summarizes the data to the programmable controller through analog-to-digital conversion. The programmable controller then uploads the data to the control host, or uploads it to the cloud through the network to form the final TBM excavation parameter data.

[0003] In recent years, the research on TBM excavation parameter data modeling, analysis and mining has received widespread attention. The proposed theoretical methods have been successfully applied in the fields of TBM excavation performance prediction, excavation parameter decision-making, surrounding rock condition identification, etc., effectively improving the excavation efficiency of TBM and providing theoretical and methodological support for TBM intelligent excavation construction.

[0004] However, TBM is fully open to the construction site, and the equipment is directly exposed to the construction site. The strong vibration and load impact generated by the excavation process, as well as the high geothermal, high humidity, and high dust working environment, lead to interference in the sensor data collection and signal transmission process. Therefore, abnormal data in the excavation parameter data is inevitable. The existence of abnormal data in TBM excavation parameters will lead to deviations in subsequent data modeling, analysis, and mining results, and even completely opposite prediction results.

[0005] There are many TBM excavation parameters. Traditional abnormal data identification methods based on data distance, spatial density and other metrics have the problem of "dimensionality disaster", which makes it difficult to play an effective role in the identification of abnormal TBM excavation parameter data. In addition, the abnormal TBM excavation parameter data is more manifested as abnormal correlation between excavation parameters. Summary of the invention

[0006] In view of the above technical problems, the present invention provides a method and system for identifying abnormal data of TBM excavation parameters. The nonlinear correlation between excavation parameters is characterized by polynomial chaos expansion (PCE), and a membership function is assigned to the excavation parameter data based on data clustering. Whether the excavation parameter data is abnormal is judged by the membership, thereby realizing effective identification of abnormal data of TBM excavation parameters.

[0007] The present invention adopts the following technical solution: a method for identifying abnormal data of TBM excavation parameters, comprising the following steps: Obtaining raw data of TBM excavation parameters, and converting the raw data into excavation parameter data in a tabular form based on message writing rules; Specifying input excavation parameters and output excavation parameters in the excavation parameter data, and constructing a polynomial chaos expansion regression model; Given the number of clustering targets, a data clustering target function is constructed in combination with the polynomial chaos expansion regression model; The Lagrange multiplier method is used to optimize and solve the data clustering objective function to obtain the membership information of the tunneling parameter data; According to the membership information, the abnormal data in the excavation parameter data is calibrated in combination with the membership information determination criterion.

[0008] In a further embodiment, the table form of the excavation parameter data is specifically expressed as follows: each row represents a piece of data, and each column represents a parameter; The conversion process of the excavation parameter data in the tabular form is as follows: Through text parsing, the original data is saved in the following data format: ;in, is the root tag, Indicates the parameter name tag, is the parameter name, Indicates parameter data tag, is the parameter value; All the spacing symbols in the data format are converted into spaces, and the excavation parameter data are read and sorted according to the table format to obtain the excavation parameter data.

[0009] In a further embodiment, the construction process of the polynomial chaos expansion regression model is as follows: Step 3.1: Based on the specified input excavation parameters and output excavation parameters, a polynomial chaos expansion regression model is established using polynomial chaos expansion. ; Step 3.2: Calculate the polynomial chaos expansion regression model by minimizing the error between the model response and the polynomial chaos expansion estimate The polynomial chaos expansion coefficients of .

[0010] In a further embodiment, the process of constructing the data clustering objective function includes: Define the given number of clustering targets as c, and the data clustering objective function is expressed using the following formula: ; In the formula, represents the data clustering objective function, represents the membership matrix, is the sample size, Indicates data points, Represents data points The true value of the output excavation parameter, Indicates The polynomial chaos expansion regression model of subclasses, Indicates Data points for each subclass The predicted value of Represents data points For The membership of the subclass, For the membership of Power operation.

[0011] In a further embodiment, the optimization solution process of the data clustering objective function is as follows: Step 5.1: Given training data , membership threshold and the number of cluster targets c ; Step 5.2: Randomly generate the initial membership matrix , in the initial membership matrix Select the membership greater than the membership threshold As the actual training data of the polynomial chaos expansion regression model, the actual training data is divided into training data subsets corresponding to the subclasses according to the clustering target number c; Step 5.3, test each training data subset to determine whether it is an empty set: if it is an empty set, repeat step 5.2 to randomly generate a new membership matrix until each training data subset is not an empty set, and execute step 5.4; Step 5.4: Construct the first Polynomial Chaos Expansion Regression Model with Subclasses ; Step 5.5: Use the Lagrange multiplier method to calculate and update the membership degree. The specific formula is as follows: ; In the formula, Represents data points credibility, , Data points representing the number of cluster targets for the tth cluster The predicted value of Step 5.6: Based on the membership updated in step 5.5, the current membership matrix is ​​further updated until the iteration termination condition is met to obtain the final membership matrix. , which is the membership information.

[0012] In a further embodiment, the abnormal data is calibrated as follows: Based on the membership information, the data points are calculated The sum of the membership degrees is expressed as ,like , it means that the corresponding data point is abnormal data, is the judgment threshold.

[0013] In a further embodiment, the polynomial chaos expansion regression model in step 3.1 The establishment process is as follows: Define the input excavation parameters as , the output excavation parameters are , based on the input excavation parameters and output excavation parameters The correlation between them is used to construct the polynomial chaos expansion regression formula, which is expressed as follows: ; In the formula, represents the polynomial chaos expansion regression model, represents the order of the polynomial chaos expansion, is a polynomial and belongs to ,in, Indicates the total order Not exceeding the given p polynomial of order, To enter the dimensions of the excavation parameters, yes The symbol for the set of dimensional non-negative integers, is the polynomial chaos expansion coefficient to be determined, To enter the excavation parameters The tensor product of univariate orthogonal polynomials corresponding to the polynomial chaotic expansion.

[0014] In a further embodiment, the calculation process of the polynomial chaos expansion coefficient in step 3.2 includes: Construct the matrix form relationship: ; In the formula, , is the coefficient vector of the polynomial chaos expansion to be determined, represents a real vector space of dimension P, P Indicates input excavation parameters The maximum order of ; , Indicates output excavation parameters The training sample vector of represents a real vector space of dimension N; , is a matrix, each column of which contains the estimated values ​​of the polynomial chaotic expansion of N samples; when When the least squares regression is used to calculate the polynomial chaos expansion coefficient vector : ; In the formula, is the permutation of the matrix; when When , the calculation problem of the polynomial chaos expansion coefficient is converted into a ℓ1 minimization problem: ; In the formula, represents the ℓ1 norm, represents the ℓ2 norm, is the truncation error of the polynomial chaotic expansion.

[0015] In a further embodiment, the input excavation parameters The maximum order of P The calculation formula is as follows: ; In the formula, is the dimension of the input tunneling parameter vector, Represents the order of the polynomial chaos expansion.

[0016] A TBM excavation parameter abnormal data identification system, used to implement the TBM excavation parameter abnormal data identification method as described above, comprising: A TBM excavation parameter processing module is configured to obtain raw data of TBM excavation parameters and convert the raw data into excavation parameter data in a tabular form based on a message writing rule; A polynomial chaos expansion regression model building module is configured to specify input excavation parameters and output excavation parameters in the excavation parameter data to build a polynomial chaos expansion regression model; A data clustering objective function construction module is set to construct a data clustering objective function in combination with the polynomial chaos expansion regression model with a given clustering target number; The data clustering objective function optimization and solving module is configured to optimize and solve the data clustering objective function based on the data clustering objective function by using the Lagrange multiplier method to obtain the membership information of the tunneling parameter data; The abnormal data identification module is configured to calibrate the abnormal data in the excavation parameter data according to the membership information and in combination with the membership information determination criteria.

[0017] The method of the invention adopts polynomial chaos expansion to characterize the correlation between TBM excavation parameters, and utilizes polynomial chaos expansion and data clustering to identify abnormal data of TBM excavation parameters, thereby solving the problem that abnormal TBM excavation parameter data is difficult to identify, and providing data support for TBM excavation parameter data modeling, analysis and mining. Compared with traditional linear models, the method has better nonlinear characterization ability; compared with black box models in the field of machine learning, it can provide reasonable interpretable results; compared with traditional metrics such as spatial distance and density, it can effectively avoid the problem of "dimensionality disaster".

[0018] The present invention constructs a data clustering objective function, establishes a clustering objective function solution strategy, obtains the TBM excavation parameter data membership, and determines whether the excavation parameter data is abnormal. Compared with the traditional spectrum-based abnormal data identification method, the present invention can adapt to the low sampling rate of TBM excavation parameter data and provide a more reasonable data division result.

[0019] The present invention realizes effective identification of abnormal data of TBM excavation parameters through a TBM excavation parameter processing module, a polynomial chaos expansion regression model construction module, a data clustering objective function construction module, a data clustering objective function optimization solution module and an abnormal data identification module. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a flow chart of the method for identifying abnormal data of TBM excavation parameters in Example 1.

[0021] Figure 2 (a) is the original data diagram of TBM excavation parameters in Example 1.

[0022] Figure 2 (b) is a diagram showing the data format after text analysis in Example 1.

[0023] Figure 3 This is the polynomial chaos expansion regression model diagram of Example 1.

[0024] Figure 4 This is a result diagram of the final excavation parameter data membership information of Example 1.

[0025] Figure 5 This is a diagram of the abnormal data identification results of TBM excavation parameters in Example 1.

[0026] Figure 6 This is an architecture diagram of a TBM excavation parameter abnormal data identification system in Example 2. DETAILED DESCRIPTION

[0027] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0028] Example 1 like Figure 1 As shown, this embodiment discloses a method for identifying abnormal data of TBM excavation parameters, comprising the following steps: The original data of TBM excavation parameters are obtained, and the original data are converted into excavation parameter data in a tabular form based on the message writing rules. In a further embodiment, the original data of TBM excavation parameters can be obtained through the data monitoring platform carried by the TBM itself, or the original data of TBM excavation parameters can be retrieved through the TBM cloud-edge collaborative monitoring platform. Correspondingly, the original data in this embodiment includes at least 2000 sample data, such as data of 460 parameters such as cutter head speed, cutter head thrust, cutter head torque, propulsion system oil pressure, temperature, motor current, etc., including 200 abnormal data, such as Figure 2 As shown in (a), the original data is in message format. Therefore, the data in message format is converted into a table, which is specifically expressed as follows: each row represents a piece of data, and each column represents a parameter, such as Figure 2 (b) shown.

[0029] In the excavation parameter data, the input excavation parameter and the output excavation parameter are specified to construct a polynomial chaos expansion regression model. In this embodiment, the advancement speed is specified as the output excavation parameter, and the excavation parameters other than the advancement speed are specified as the input excavation parameters. Based on this, a polynomial chaos expansion regression model is constructed. Figure 3 .

[0030] Given the number of clustering targets, the data clustering objective function is constructed in combination with the polynomial chaos expansion regression model. That is, the clustering objective function is constructed based on the error between the estimated value and the true value of the polynomial chaos expansion regression model of the advancing speed to evaluate the current TBM excavation parameter data clustering results.

[0031] Based on the data clustering objective function, the Lagrange multiplier method is used to optimize and solve the data clustering objective function to obtain the membership information of the tunneling parameter data; in other words, the clustering objective function is updated through alternating iterations of membership and data attribution information, and the iteration is terminated according to the iteration termination condition to obtain the final membership information of the tunneling parameter data, such as Figure 4 shown.

[0032] According to the membership information, the abnormal data in the excavation parameter data is calibrated in combination with the membership information judgment criteria, such as Figure 5 shown.

[0033] In a further embodiment, according to the above description, the message format is xml The process of converting the excavation parameter data into a table format is as follows: Through text parsing, the original data is saved in the following data format: ; in, is the root tag, Indicates the parameter name tag, is the parameter name, Indicates parameter data tag, is the parameter value; use txt The editing file converts all the spacing symbols in the data format into spaces, and reads and organizes the tunneling parameter data in the form of the table. It is further understood that the message data of each line in the message is read in turn, and the corresponding conversion is performed to finally obtain the data of the entire file.

[0034] In a further embodiment, the construction process of the polynomial chaos expansion regression model is as follows: Step 3.1: Based on the specified input excavation parameters and output excavation parameters, a polynomial chaos expansion regression model is established using polynomial chaos expansion. ; Step 3.2: Calculate the polynomial chaos expansion regression model by minimizing the error between the model response and the polynomial chaos expansion estimate The polynomial chaos expansion coefficients of .

[0035] Furthermore, the polynomial chaos expansion regression model in step 3.1 The establishment process is as follows: Define the input excavation parameters as , the output excavation parameters are , based on the input excavation parameters and output excavation parameters The correlation between them is used to construct the polynomial chaos expansion regression formula, which is expressed as follows: ; In the formula, represents the polynomial chaos expansion regression model, represents the order of the polynomial chaos expansion, is a polynomial and belongs to ,in, Indicates the total order Not exceeding the given p polynomial of order, To enter the dimensions of the excavation parameters, yes The symbol for the set of dimensional non-negative integers, is the polynomial chaos expansion coefficient to be determined, To enter the excavation parameters The tensor product of univariate orthogonal polynomials corresponding to the polynomial chaotic expansion.

[0036] Correspondingly, the calculation process of the polynomial chaos expansion coefficient in step 3.2 includes: Construct the matrix form relationship: ; In the formula, , is the coefficient vector of the polynomial chaos expansion to be determined, represents a real vector space of dimension P, P Indicates input excavation parameters The maximum order of ; , Indicates output excavation parameters The training sample vector of represents a real vector space of dimension N; , is a matrix, each column of which contains the estimated values ​​of the polynomial chaotic expansion of N samples; when When the least squares regression is used to calculate the polynomial chaos expansion coefficient vector : ; In the formula, is the permutation of the matrix; when When , the calculation problem of the polynomial chaos expansion coefficient is converted into a ℓ1 minimization problem: ; In the formula, represents the ℓ1 norm, represents the ℓ2 norm, is the truncation error of the polynomial chaotic expansion.

[0037] It is worth mentioning that the input excavation parameters The maximum order of P The calculation formula is as follows: ; In the formula, is the dimension of the input excavation parameter vector.

[0038] In another embodiment, the process of constructing the data clustering objective function includes: Define the given number of clustering targets as c. In this embodiment, the value of c is 3. The data clustering objective function is expressed by the following formula: ; In the formula, represents the data clustering objective function, represents the membership matrix, is the sample size, Indicates data points, Represents data points The true value of the output excavation parameter, Indicates The polynomial chaos expansion regression model of subclasses, Indicates Data points for each subclass The predicted value of Represents data points For The membership of the subclass, For the membership of Power operation.

[0039] Among them, the data points For The membership of subclass The following conditions should be met: ;in, Indicates any For data points The credibility is calculated as follows: ; in, Represents the true value The minimum absolute error between the predicted value and the corresponding It represents the minimum absolute error between the true value and the predicted value corresponding to any data point j.

[0040] It should be noted that ;in, Represents data points The true value of the output excavation parameter, Indicates Polynomial Chaos Expansion Regression Model with Subclasses For data points The predicted values ​​of the output excavation parameters are: Polynomial Chaos Expansion Regression Model with Subclasses The method of obtaining can be polynomial chaos expansion regression model A description of how to obtain the .

[0041] Based on the given data clustering objective function , the Lagrange multiplier method is used to solve the clustering objective function and obtain the membership information of the tunneling parameter data, including the following steps: Step 5.1: Given training data , membership threshold and the number of cluster targets c ; Step 5.2: Randomly generate the initial membership matrix , in the initial membership matrix Select the membership greater than the membership threshold As the actual training data of the polynomial chaos expansion regression model, the actual training data is divided into training data subsets corresponding to the subclasses according to the clustering target number c; Step 5.3, test each training data subset to determine whether it is an empty set: if it is an empty set, repeat step 5.2 to randomly generate a new membership matrix until each training data subset is not an empty set, and execute step 5.4; Step 5.4: Construct the first Polynomial Chaos Expansion Regression Model with Subclasses ; Step 5.5: Use the Lagrange multiplier method to calculate and update the membership degree. The specific formula is as follows: ; In the formula, Represents data points credibility, , Data points representing the number of cluster targets for the tth cluster The predicted value of .

[0042] Step 5.6: Based on the membership updated in step 5.5, the current membership matrix is ​​further updated until the iteration termination condition is met to obtain the final membership matrix. , which is the membership information. The iteration termination condition may be a set maximum number of iterations, such as 100 iterations. In other embodiments, other iteration termination conditions may also be used.

[0043] Based on the above description, the calibration method of abnormal data in this embodiment is as follows: Based on the membership information, the data points are calculated The sum of the membership degrees is expressed as ,like , it means that the corresponding data point is abnormal data, is the judgment threshold. The value of is 0.8, combined with Figure 4 and 5 It can be seen that there are significant differences in the degree of membership between abnormal data and normal data. Figure 5 The label 1 indicates an abnormality. It can be seen that the method for identifying abnormal data of TBM excavation parameters provided in this embodiment can accurately identify the abnormal data of TBM excavation parameters in this case.

[0044] In summary, based on the unprocessed TBM excavation parameter data and the TBM excavation parameter data with abnormal data removed, the polynomial chaos expansion is used to establish the advancement speed prediction model, and it is found that the prediction accuracy is significantly improved. This embodiment solves the problem that abnormal data of TBM excavation parameters is difficult to identify, and provides a reference for TBM state identification, fault diagnosis, subsequent data modeling, analysis and mining.

[0045] Example 2 In order to implement the method for identifying abnormal data of TBM excavation parameters described in Example 1, this embodiment discloses a system for identifying abnormal data of TBM excavation parameters, such as Figure 6 Shown include: A TBM excavation parameter processing module is configured to obtain raw data of TBM excavation parameters and convert the raw data into excavation parameter data in a tabular form based on a message writing rule; A polynomial chaos expansion regression model building module is configured to specify input excavation parameters and output excavation parameters in the excavation parameter data to build a polynomial chaos expansion regression model; A data clustering objective function construction module is set to construct a data clustering objective function in combination with the polynomial chaos expansion regression model with a given clustering target number; The data clustering objective function optimization and solving module is configured to optimize and solve the data clustering objective function based on the data clustering objective function by using the Lagrange multiplier method to obtain the membership information of the tunneling parameter data; The abnormal data identification module is configured to calibrate the abnormal data in the excavation parameter data according to the membership information and in combination with the membership information determination criteria.

[0046] It should be understood that although this specification is described according to various embodiments, not every embodiment contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

[0047] The series of detailed descriptions listed above are only specific descriptions of feasible embodiments of the present invention. They are not intended to limit the scope of protection of the present invention. All equivalent embodiments or changes that do not deviate from the technical spirit of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for identifying abnormal data of TBM excavation parameters, characterized in that: The following steps are involved: Acquire the original data of TBM excavation parameters, and convert the original data into excavation parameter data in a tabular form based on the message writing rules; Specifying input excavation parameters and output excavation parameters in the excavation parameter data, and constructing a polynomial chaos expansion regression model; Given the number of clustering targets, a data clustering target function is constructed in combination with the polynomial chaos expansion regression model; The Lagrange multiplier method is used to optimize and solve the data clustering objective function to obtain the membership information of the tunneling parameter data; According to the membership information and in combination with the membership information determination criteria, abnormal data in the excavation parameter data are calibrated.

2. A method for identifying abnormal data of TBM excavation parameters according to claim 1, characterized in that: The table form of the excavation parameter data is specifically expressed as follows: each row represents a piece of data, and each column represents a parameter; The conversion process of the excavation parameter data in the tabular form is as follows: Through text parsing, the original data is saved in the following data format: ;in, is the root tag, Indicates the parameter name tag, is the parameter name, Indicates parameter data tag, is the parameter value; All the spacing symbols in the data format are converted into spaces, and the excavation parameter data are read and sorted according to the table format to obtain the excavation parameter data.

3. The method for identifying abnormal data of TBM excavation parameters according to claim 1, characterized in that: The construction process of the polynomial chaos expansion regression model is as follows: Step 3.1: Based on the specified input excavation parameters and output excavation parameters, a polynomial chaos expansion regression model is established using polynomial chaos expansion. ; Step 3.2: Calculate the polynomial chaos expansion regression model by minimizing the error between the model response and the polynomial chaos expansion estimate The polynomial chaos expansion coefficients of .

4. A method for identifying abnormal data of TBM excavation parameters according to claim 1, characterized in that: The construction process of the data clustering objective function includes: Define the given number of clustering targets as c , the data clustering objective function is expressed using the following formula: ； In the formula, represents the data clustering objective function, represents the membership matrix, is the sample size, Indicates data points, Represents data points The true value of the output excavation parameter, Indicates The polynomial chaos expansion regression model of subclasses, Indicates Data points for each subclass The predicted value of Represents data points For The membership of the subclass, For the membership of Power operation.

5. A method for identifying abnormal data of TBM excavation parameters according to claim 4, characterized in that: The optimization process of the data clustering objective function is as follows: Step 5.1: Given training data , membership threshold and the number of cluster targets c ; Step 5.2: Randomly generate the initial membership matrix , in the initial membership matrix Select the membership greater than the membership threshold As the actual training data of the polynomial chaos expansion regression model, the actual training data is divided into training data subsets corresponding to the subclasses according to the clustering target number c; Step 5.3, test each training data subset to determine whether it is an empty set: if it is an empty set, repeat step 5.2 to randomly generate a new membership matrix until each training data subset is not an empty set, and execute step 5.4; Step 5.4: Construct the first Polynomial Chaos Expansion Regression Model with Subclasses ; Step 5.5: Use the Lagrange multiplier method to calculate and update the membership degree. The specific formula is as follows: ； In the formula, Represents data points credibility, , Data points representing the number of cluster targets for the tth cluster The predicted value of Step 5.6: Based on the membership updated in step 5.5, the current membership matrix is ​​further updated until the iteration termination condition is met to obtain the final membership matrix. , which is the membership information.

6. A method for identifying abnormal data of TBM excavation parameters according to claim 1, characterized in that: The calibration method of the abnormal data is as follows: Based on the membership information, the data points are calculated The sum of the membership degrees is expressed as ,like , it means that the corresponding data point is abnormal data, is the judgment threshold.

7. A method for identifying abnormal data of TBM excavation parameters according to claim 3, characterized in that: The polynomial chaos expansion regression model in step 3.1 The establishment process is as follows: Define the input excavation parameters as , the output excavation parameters are , based on the input excavation parameters and output excavation parameters The correlation between them is used to construct the polynomial chaos expansion regression formula, which is expressed as follows: ； In the formula, represents the polynomial chaos expansion regression model, represents the order of the polynomial chaos expansion, is a polynomial and belongs to ,in, Indicates the total order Not exceeding the given p polynomial of order, To enter the dimensions of the excavation parameters, yes The symbol for the set of dimensional non-negative integers, is the polynomial chaos expansion coefficient to be determined, To enter the excavation parameters The tensor product of univariate orthogonal polynomials corresponding to the polynomial chaotic expansion.

8. The method for identifying abnormal data of TBM excavation parameters according to claim 3, characterized in that: The calculation process of the polynomial chaos expansion coefficient in step 3.2 includes: Construct the matrix form relationship: ; In the formula, , is the coefficient vector of the polynomial chaos expansion to be determined, represents a real vector space of dimension P, P Indicates input excavation parameters The maximum order of ; , Indicates output excavation parameters The training sample vector of represents a real vector space of dimension N; , is a matrix, each column of which contains the estimated values ​​of the polynomial chaotic expansion of N samples; when When the least squares regression is used to calculate the polynomial chaos expansion coefficient vector : ; In the formula, is the permutation of the matrix; when When , the calculation problem of the polynomial chaos expansion coefficient is converted into a ℓ1 minimization problem: ; In the formula, represents the ℓ1 norm, represents the ℓ2 norm, is the truncation error of the polynomial chaotic expansion.

9. A method for identifying abnormal data of TBM excavation parameters according to claim 8, characterized in that: The input excavation parameters The maximum order of P The calculation formula is as follows: ； In the formula, is the dimension of the input tunneling parameter vector, Represents the order of the polynomial chaos expansion.

10. A TBM excavation parameter abnormal data identification system, used to implement the TBM excavation parameter abnormal data identification method according to any one of claims 1 to 9, characterized in that: include: A TBM excavation parameter processing module is configured to obtain raw data of TBM excavation parameters and convert the raw data into excavation parameter data in a tabular form based on a message writing rule; A polynomial chaos expansion regression model building module is configured to specify input excavation parameters and output excavation parameters in the excavation parameter data to build a polynomial chaos expansion regression model; A data clustering objective function construction module is set to construct a data clustering objective function in combination with the polynomial chaos expansion regression model with a given clustering target number; The data clustering objective function optimization and solving module is configured to optimize and solve the data clustering objective function based on the data clustering objective function by using the Lagrange multiplier method to obtain the membership information of the tunneling parameter data; The abnormal data identification module is configured to calibrate the abnormal data in the excavation parameter data according to the membership information and in combination with the membership information determination criteria.

Images