Shield abnormal data detection method and system based on Z-score model

Abstract

The present invention discloses a shield abnormal data detection method and system based on a Z-score model, which relates to the technical field of shield construction, including: using intelligent means to collect shield data, and uploading the shield data to the cloud for storage; analyzing and processing the collected shield data to obtain pre-processed shield data; performing data correlation analysis on the pre-processed shield data to confirm parameters that have an impact on the overall data; selecting basic attributes of shield data, and performing abnormal data detection based on the Z-score model. The present invention can effectively carry out shield abnormal data early warning and improve the intelligent level of shield construction engineering.

Description

Technical Field

[0001] The present invention relates to the technical field of shield construction, and more specifically to a shield abnormal data detection method and system based on a Z-score model. Background Art

[0002] During shield construction, in order to ensure the safe operation of the shield machine, geological surveys are usually conducted on the excavation route to determine the geological conditions and identify abnormal geological conditions in advance to reduce the risks during shield construction. However, since the shield machine is affected by the imbalance of soil pressure on the excavation surface and the outer surface of the shield shell, changes in underground soil layers and other factors during its operation, it is impossible to accurately provide early warnings of abnormal working conditions based only on the information from advance surveys. In order to ensure construction safety, it is necessary to analyze and study the data during shield construction to achieve early warnings of abnormal working conditions.

[0003] The traditional abnormal working condition early warning method is to manually record the relevant parameter values. When the deviation exceeds a certain range, it is reported to the superior department for processing. This method analyzes and studies the original data and cannot intuitively see the degree of deviation of each data. Therefore, how to solve the subjectivity and deviation problems of manual data screening and effectively warn of abnormal shield data is a technical problem that technicians in this field urgently need to solve. Summary of the invention

[0004] In view of this, the present invention provides a shield abnormal data detection method and system based on the Z-score model, which can effectively perform shield abnormal data early warning and improve the intelligence level of shield construction projects.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] A shield abnormal data detection method based on a Z-score model comprises the following steps:

[0007] Collect shield data using intelligent means, and upload the shield data to the cloud for storage;

[0008] Analyzing and processing the collected shield data to obtain pre-processed shield data;

[0009] Performing data correlation analysis on the pre-processed shield data to confirm parameters that have an impact on the overall data;

[0010] The basic attributes of shield data are selected, and abnormal data detection is performed based on the Z-score model.

[0011] The technical effects achieved by the above technology are: introducing Z-score to measure the degree of data deviation, using data analysis methods to screen data to address the subjectivity and bias problems of manual data screening, and performing abnormal data detection and early warning based on the proposed Z-score method.

[0012] Optionally, the specific method of collecting and storing the shield data is:

[0013] The PLC data of the shield machine during its operation is collected through the SCADA system, and the communication module is used to transmit the PLC data to the control console in real time, so as to achieve the goal of data collection and monitoring;

[0014] Data is extracted from the PLC data transmitted to the console according to parameter requirements and stored in a SQL database for subsequent data analysis and storage of the calculation process.

[0015] Optionally, the obtaining of pre-processed shield data specifically includes the following steps:

[0016] Using a clustering algorithm in data mining, abnormal data is identified and removed from the shield data;

[0017] Eliminate the data samples with missing data and obtain the preprocessed shield data.

[0018] The technical effect achieved by the above technical scheme is: it provides a method for preprocessing shield data, which can eliminate certain data that deviate greatly from normal data from the shield data to ensure the validity of the data, and can also eliminate data samples that have no practical significance due to missing parameters in the data.

[0019] Optionally, the data correlation analysis is specifically performed as follows:

[0020] The pre-processed shield data is used as training data, and a correlation analysis algorithm is used to perform correlation parameter analysis on all parameters;

[0021] Based on the correlation analysis results and the construction project experience parameters, a comprehensive analysis is performed to obtain the parameters that have an impact on the overall data, which are used as adjustment parameters for subsequent alarm release plan recommendations.

[0022] Optionally, the abnormal data detection specifically includes the following steps:

[0023] The basic attributes of shield data that can characterize abnormal conditions during the operation of the shield machine are selected, namely, total thrust, propulsion speed, cutter head torque, and cutter head speed;

[0024] The definition of the Z-score model is as follows:

[0025]

[0026]

[0027]

[0028] Where: x i represents the observed value, μ represents the population mean, and δ represents the population standard deviation;

[0029] Assuming that the parameter eigenvalues ​​follow a normal distribution, according to the basic principles of mathematical statistics, the range of the outlier value of the Z-score is determined as:

[0030] |Z-score|>3.

[0031] Optionally, the method further includes: when the Z-score is within the interval [2,3], a reminder warning is performed, specifically:

[0032] When 2<|Z-score|≤2.7, a first-level warning signal is issued;

[0033] When 2.7<|Z-score|≤3, a second-level warning signal is issued.

[0034] The present invention also provides a shield abnormal data detection system based on the Z-score model, comprising: an acquisition module, a storage module, a preprocessing module, an analysis module, and an abnormality detection module;

[0035] The acquisition module is used to collect shield data through intelligent means;

[0036] The storage module is used to upload the shield data to the cloud for storage;

[0037] The preprocessing module is used to analyze and process the collected shield data to obtain preprocessed shield data;

[0038] The analysis module is used to perform data correlation analysis on the pre-processed shield data to confirm the parameters that have an impact on the overall data;

[0039] The anomaly detection module is used to select basic attributes of shield data and perform anomaly data detection based on the Z-score model.

[0040] Optionally, the preprocessing module includes an abnormal data processing submodule and a missing data processing submodule;

[0041] The abnormal data processing submodule identifies and removes abnormal data from the shield data through a clustering algorithm in data mining;

[0042] The missing data processing submodule is used to remove data samples with missing data.

[0043] Optionally, the basic attributes of the shield data include total thrust, propulsion speed, cutter head torque, and cutter head rotation speed.

[0044] It can be seen from the above technical scheme that compared with the prior art, the present invention discloses a shield abnormal data detection method and system based on the Z-score model, introduces the Z-score abnormal detection method to measure the degree of data deviation, and uses data analysis methods to screen abnormal data in view of the subjectivity and bias problems of manual data screening. Abnormal data detection and early warning are performed based on the proposed Z-score, which can effectively improve the intelligent information level of shield construction projects and enhance the technical strength of enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0046] Figure 1 It is a flow chart of the shield abnormal data detection method based on the Z-score model;

[0047] Figure 2 This is the structural diagram of the shield abnormal data detection system based on the Z-score model. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] Example 1

[0050] During shield construction, the shield machine can collect various data during the shield excavation process. In engineering, the abnormal working condition of the shield is usually warned by setting thresholds for the main parameters. However, relying solely on this method, the linkage between the parameters is ignored, resulting in missed reports or false reports. By defining the excavation index by calculating the main parameters, the abnormal working condition of the shield can be warned more effectively.

[0051] Therefore, the embodiment of the present invention discloses a shield abnormal data detection method based on the Z-score model, comprising the following steps:

[0052] Use intelligent means to collect shield data and upload the shield data to the cloud for storage;

[0053] Analyze and process the collected shield data to obtain pre-processed shield data;

[0054] Conduct data correlation analysis on pre-processed shield data to identify parameters that affect the overall data;

[0055] The basic attributes of shield data are selected, and abnormal data detection is performed based on the Z-score model.

[0056] Next, refer to Figure 1 , the abnormal working condition analysis based on the tunneling index in this technical solution is elaborated in detail.

[0057] (1) Shield data collection

[0058] The shield data needs to be collected in real time through intelligent means and uploaded to the cloud for storage. In this embodiment, based on the existing construction technology and environment, the PLC data of the shield machine during its movement can be collected through the SCADA system (data acquisition and monitoring control system), and the communication module can be used to transmit the PLC data to the control console in real time to achieve the goal of data collection and monitoring.

[0059] In this method, data is extracted from the PLC data transmitted to the console according to parameter requirements and stored in a sql database for subsequent data analysis and storage of process values.

[0060] (2) Data analysis and processing

[0061] 1) Abnormal data processing. There are some data in the shield tunneling data that are greatly deviated from the normal data. This part of the data may be caused by the abnormality of the acquisition system. Therefore, the clustering algorithm in data mining can be used to identify and remove abnormal data from the shield data to ensure the validity of the data;

[0062] 2) Missing data processing: If there are missing parameters in the data, then the data has no practical significance, and the data samples with missing data can be eliminated.

[0063] (3) Data correlation analysis

[0064] The preprocessed shield data is used as training data, and the correlation analysis algorithm is used to perform correlation parameter analysis on all parameters. A comprehensive analysis is performed based on the correlation analysis results and the empirical parameters of the construction project to obtain the parameters that have an impact on the overall data, which are used as adjustment parameters for subsequent alarm release plan recommendations.

[0065] (4) Z-score abnormal data detection and early warning

[0066] Combining the results of correlation analysis and expert experience, it can be seen that the basic attributes of shield tunneling data are: total thrust, propulsion speed, cutter head torque, and cutter head speed. These four can characterize the abnormal conditions during the operation of the shield machine. Therefore, these four basic attributes are selected for abnormal data detection.

[0067] The definition of the Z-score model is as follows:

[0068]

[0069]

[0070]

[0071] Where: x i represents the observed value, μ represents the population mean, and δ represents the population standard deviation;

[0072] Assuming that the parameter eigenvalues ​​follow a normal distribution, according to the basic principles of mathematical statistics, under normal circumstances, the proportion of random variables taking values ​​within the range of (μ±3δ) is 99.73%, and the probability of exceeding this range is only 0.0027, which is a small probability event and usually does not occur. Based on this, the range of the abnormal value of Z-score can be determined as:

[0073] |Z-score|>3.

[0074] When the Z-score is in the range of [2,3], a warning will be issued, specifically:

[0075] When 2<|Z-score|≤2.7, a first-level warning signal is issued;

[0076] When 2.7<|Z-score|≤3, a second-level warning signal is issued.

[0077] Example 2

[0078] This embodiment discloses a shield abnormal data detection system based on the Z-score model. Figure 2 As shown, it includes: a collection module, a storage module, a preprocessing module, an analysis module, and an anomaly detection module;

[0079] The acquisition module is used to collect shield data through intelligent means;

[0080] Storage module, used to upload shield data to the cloud for storage;

[0081] The preprocessing module is used to analyze and process the collected shield data to obtain preprocessed shield data;

[0082] The analysis module is used to perform data correlation analysis on the pre-processed shield data and confirm the parameters that affect the overall data;

[0083] The anomaly detection module is used to select the basic attributes of shield data and perform anomaly data detection based on the Z-score model.

[0084] Further, the preprocessing module includes an abnormal data processing submodule and a missing data processing submodule;

[0085] The abnormal data processing submodule identifies and removes abnormal data from shield data through clustering algorithms in data mining;

[0086] The missing data processing submodule is used to remove data samples with missing data.

[0087] Specifically, the basic attributes of the shield data include total thrust, propulsion speed, cutter head torque, and cutter head rotation speed.

[0088] The traditional method of manually warning of abnormal working conditions has problems of subjectivity and bias. This technical solution introduces the Z-score anomaly detection method to measure the degree of data deviation, uses data analysis methods to screen abnormal data, and performs abnormal data detection and warning based on the proposed Z-score, which can effectively improve the intelligent information level of shield construction projects and enhance the company's technical strength.

[0089] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0090] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A shield abnormal data detection method based on Z-score model, characterized in that: The following steps are involved: Collect shield data using intelligent means, and upload the shield data to the cloud for storage; Analyze and process the collected shield data to obtain pre-processed shield data; specifically: use a clustering algorithm in data mining to identify and remove abnormal data from the shield data; remove data samples with missing data to obtain pre-processed shield data; Performing data correlation analysis on the pre-processed shield data to identify parameters that have an impact on the overall data; specifically: using the pre-processed shield data as training data, applying a correlation analysis algorithm, and performing correlation parameter analysis on all parameters; performing a comprehensive analysis based on the correlation analysis results and the empirical parameters of the construction project to obtain parameters that have an impact on the overall data; The basic attributes of shield data are selected, and abnormal data detection is performed based on the Z-score model.

2. According to the method for detecting abnormal data of shield machine based on Z-score model in claim 1, it is characterized in that: The specific method of collecting and storing the shield data is as follows: The PLC data of the shield machine during its operation is collected through the SCADA system, and the PLC data is transmitted to the control console in real time using the communication module; Data is extracted from the PLC data transmitted to the console according to parameter requirements and stored in a sql database.

3. According to the method for detecting abnormal data of shield machine based on Z-score model in claim 1, it is characterized in that: The abnormal data detection specifically includes the following steps: The basic attributes of shield data that can characterize abnormal conditions during the operation of the shield machine are selected, namely, total thrust, propulsion speed, cutter head torque, and cutter head speed; The Z-score model is defined as follows: Where: x i represents the observed value, μ represents the population mean, and δ represents the population standard deviation; Assuming that the parameter eigenvalues ​​follow a normal distribution, according to the basic principles of mathematical statistics, the range of the outlier value of the Z-score is determined as: |Z-score|>3.

4. According to the method for detecting abnormal data of shield machine based on Z-score model in claim 3, it is characterized in that: The method further includes: when the Z-score is within the interval [2,3], a reminder warning is performed, specifically: When 2<|Z-score|≤2.7, a first-level warning signal is issued; When 2.7<|Z-score|≤3, a second-level warning signal is issued.

5. A shield abnormal data detection system based on the Z-score model, characterized in that: include: Acquisition module, storage module, preprocessing module, analysis module, anomaly detection module; The acquisition module is used to collect shield data through intelligent means; The storage module is used to upload the shield data to the cloud for storage; The preprocessing module is used to analyze and process the collected shield data to obtain preprocessed shield data; wherein the preprocessing module includes an abnormal data processing submodule and a missing data processing submodule; the abnormal data processing submodule identifies and removes abnormal data from the shield data through a clustering algorithm in data mining; The missing data processing submodule is used to remove data samples with missing data; The analysis module is used to perform data correlation analysis on the pre-processed shield data to confirm the parameters that have an impact on the overall data; specifically, the pre-processed shield data is used as training data, and a correlation analysis algorithm is used to perform correlation parameter analysis on all parameters; a comprehensive analysis is performed based on the correlation analysis results and the construction project experience parameters to obtain the parameters that have an impact on the overall data; The anomaly detection module is used to select basic attributes of shield data and perform anomaly data detection based on the Z-score model.

6. The shield abnormal data detection system based on the Z-score model according to claim 5 is characterized in that: The basic attributes of the shield data include total thrust, propulsion speed, cutter head torque, and cutter head rotation speed.

Images