Data processing method and device for tunneling parameter prediction
Through causal analysis and preprocessing, a tunneling parameter prediction model was constructed, which solved the problem of low prediction accuracy of TBM tunneling construction parameters, and achieved higher prediction accuracy and basis for parameter adjustment.
Patent Information
- Application Number
- CN202411704871.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-26
- Publication Date
- 2025-03-28
AI Technical Summary
In the prior art, the prediction accuracy of TBM excavation construction parameters is low, and the AI model has a 'black box' problem when revealing the nonlinear relationship between input and output indicators, which is not conducive to parameter adjustment and decision-making.
A data processing method is proposed, by obtaining the data to be processed, performing feature parameters screening and preprocessing based on causal discovery, building a tunneling parameter prediction model, and using causal relationship characteristics and related data for parameter prediction.
The accuracy of TBM excavation construction parameters prediction is improved, the ‘black box’ problem is solved, and the basis for parameter adjustment and decision-making is enhanced.
Smart Images
Figure CN119719620A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of computers, and in particular to a data processing method and device for predicting tunneling parameters. Background Art
[0002] Standardization, mechanization and automation of underground engineering construction are the future development trends. In recent years, the number of tunnel projects constructed using full-section hard rock tunnel boring machines (TBMs) has gradually increased. During TBM excavation construction, the parameter setting of TBMs has a great impact on the quality of TBM construction, but TBM parameters are difficult to predict and adjust. Currently, they rely on the experience of operators, resulting in large performance randomness and difficulty in detecting abnormalities.
[0003] In the existing technology, the TBM parameter combination is analyzed and predicted by using theoretical calculations and empirical models, but the prediction accuracy is low due to many assumptions and small amount of data. With the development of AI technology, by exploring the nonlinear relationship between TBM parameters in "big data" and building AI models to characterize the nonlinear relationship between these features, it can be used for high-precision parameter prediction. The relevant research methods can be divided into two categories: machine learning and deep learning. However, these methods fail to fully utilize the advantages of TBM's operating data (time series data), and the AI model has a "black box" problem when revealing the nonlinear relationship between input and output indicators, which is not conducive to parameter adjustment and decision-making in TBM construction.
[0004] Therefore, this application is proposed to address the problems existing in the prediction of TBM excavation construction parameters. Summary of the invention
[0005] The main purpose of this application is to provide a data processing method for predicting tunneling parameters, so as to solve the problems existing in the prior art in predicting TBM tunneling construction parameters and achieve the technical effect of improving the accuracy of TBM tunneling construction parameter prediction.
[0006] In order to achieve the above-mentioned purpose, in a first aspect of the present application, a data processing method for predicting tunneling parameters is proposed, comprising:
[0007] Acquiring data to be processed, wherein the data to be processed is relevant data for representing excavation construction parameters;
[0008] Performing feature parameter screening processing based on causal discovery on the data to be processed to obtain input feature parameter data, wherein the input feature parameter data is data for representing feature parameters for predicting tunneling parameters in the data to be processed;
[0009] The input characteristic parameter data is subjected to parameter prediction processing based on a tunneling parameter prediction model to obtain predicted tunneling parameter data.
[0010] Furthermore, the data to be processed is subjected to feature parameter screening processing based on causal discovery, and the input feature parameter data obtained includes:
[0011] Performing normalized data preprocessing on the data to be processed to obtain preprocessed tunneling parameter data, wherein the preprocessed tunneling parameter data is used to represent the data to be processed after normalization processing;
[0012] Performing causal relationship strength analysis processing based on causal discovery on the pre-processed excavation parameter data to obtain parameter causal relationship strength characteristic data, wherein the parameter causal relationship strength characteristic data is characteristic data used to represent the causal relationship strength of characteristic parameters;
[0013] The parameter causal relationship strength characteristic data is screened based on a preset causal relationship strength threshold to obtain the input characteristic parameter data.
[0014] Furthermore, the parameter causal relationship strength characteristic data is screened based on a preset causal relationship strength threshold, and the input characteristic parameter data obtained includes:
[0015] Comparing the parameter causal relationship strength characteristic data with the preset causal relationship strength threshold to implement screening processing of characteristic parameters,
[0016] If the strength value of the parameter causal relationship strength feature data is greater than the preset causal relationship strength threshold, obtaining input feature data;
[0017] Parameter data corresponding to the input characteristic data are matched in the pre-processing excavation parameters to obtain the input characteristic parameter data.
[0018] Furthermore, the input characteristic parameter data is subjected to parameter prediction processing based on the tunneling parameter prediction model to obtain predicted tunneling parameter data including:
[0019] Performing causal feature extraction processing based on causal discovery on the input feature parameter data to obtain input causal feature data, wherein the input causal feature data is feature data used to represent the causal relationship in the input feature parameters;
[0020] Performing data splicing processing on the input causal feature data and the input feature parameter data to obtain spliced input parameter data;
[0021] The spliced data parameter data is subjected to parameter prediction processing based on a tunneling parameter prediction model to obtain the predicted tunneling parameter data.
[0022] Furthermore, before obtaining the data to be processed, the method further includes:
[0023] Acquire sample data to be processed, wherein the sample data to be processed is excavation parameter sample data used to train the excavation parameter prediction model;
[0024] Performing feature parameter screening processing based on causal discovery on the sample data to be processed to obtain sample input feature parameter data, wherein the sample input feature parameter data is data used to represent input feature parameters of the model to be trained in the sample data to be processed;
[0025] The preset deep learning model is trained according to the sample input characteristic parameter data and the sample data to be processed to obtain the tunneling parameter prediction model.
[0026] Furthermore, the preset deep learning model is trained according to the sample input characteristic parameter data and the sample data to be processed to obtain the tunneling parameter prediction model, which includes:
[0027] Performing causal feature extraction processing based on causal discovery on the sample input feature parameter data to obtain sample input causal feature data, wherein the sample input causal feature data is feature data used to represent the causal relationship in the sample input feature parameters;
[0028] Performing data splicing processing on the sample input causal feature data and the sample input feature parameter data to obtain sample splicing input parameter data;
[0029] Performing output characteristic parameter extraction processing on the sample data to be processed to obtain sample output characteristic parameter data;
[0030] The sample splicing input parameter data is used as the input of the preset deep learning model, and the sample output feature parameter data is used as the output of the preset deep learning model. The preset deep learning model is subjected to model training processing to obtain the excavation parameter prediction model.
[0031] Furthermore, the sample splicing input parameter data is used as the input of the preset deep learning model, the sample output characteristic parameter data is used as the output of the preset deep learning model, and the preset deep learning model is subjected to model training processing to obtain the tunneling parameter prediction model, which includes:
[0032] The sample splicing input parameter data is used as the input of the preset deep learning model, and the sample output characteristic parameter data is used as the output of the preset deep learning model, and the preset deep learning model is subjected to model training processing to obtain a process excavation parameter prediction model;
[0033] Performing model evaluation processing on the process excavation parameter prediction model to obtain model evaluation characteristic data, wherein the model evaluation characteristic data is characteristic data used to represent the performance of the process excavation parameter prediction model;
[0034] The model evaluation feature data is judged based on preset performance evaluation rules.
[0035] If the model evaluation characteristic data satisfies the preset performance evaluation rules, the excavation parameter prediction model is obtained, wherein the excavation parameter prediction model is a process excavation parameter prediction model that satisfies the preset performance evaluation rules;
[0036] If the model evaluation feature data does not satisfy the preset performance evaluation rules, the process excavation parameter prediction model is subjected to model iteration training processing until the process excavation parameter prediction model satisfies the preset performance evaluation rules, thereby obtaining the excavation parameter prediction model.
[0037] According to a second aspect of the present application, a data processing device for predicting tunneling parameters is proposed, comprising:
[0038] A data acquisition module, used to acquire data to be processed, wherein the data to be processed is relevant data used to represent excavation construction parameters;
[0039] A characteristic parameter screening module is used to perform characteristic parameter screening processing on the data to be processed based on causal discovery to obtain input characteristic parameter data, wherein the input characteristic parameter data is data for representing characteristic parameters for predicting tunneling parameters in the data to be processed;
[0040] The excavation parameter prediction module is used to perform parameter prediction processing on the input characteristic parameter data based on the excavation parameter prediction model to obtain predicted excavation parameter data.
[0041] According to a third aspect of the present application, a computer-readable storage medium is proposed, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the above-mentioned data processing method for tunneling parameter prediction.
[0042] According to the fourth aspect of the present application, an electronic device is proposed, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor executes the above-mentioned data processing method for predicting excavation parameters.
[0043] The technical solution provided by the embodiments of the present application may have the following beneficial effects:
[0044] In the present application, data to be processed is obtained, wherein the data to be processed is related data for representing excavation construction parameters; the data to be processed is subjected to characteristic parameter screening processing based on causal discovery to obtain input characteristic parameter data, wherein the input characteristic parameter data is data for representing characteristic parameters for predicting excavation parameters in the data to be processed; the input characteristic parameter data is subjected to parameter prediction processing based on an excavation parameter prediction model to obtain predicted excavation parameter data. By performing causal analysis on excavation construction parameters, the excavation parameter prediction model performs parameter prediction processing on causal relationship characteristics and related excavation parameter data, thereby solving the problem of low accuracy of TBM excavation parameter prediction in the prior art and achieving the technical effect of improving the accuracy of excavation parameter prediction. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] The drawings constituting a part of this application are used to provide a further understanding of this application, so that other features, purposes and advantages of this application become more obvious. The schematic embodiment drawings and their descriptions of this application are used to explain this application and do not constitute an improper limitation on this application. In the drawings:
[0046] Figure 1 A flow chart of a data processing method for predicting tunneling parameters provided in this application;
[0047] Figure 2 A flow chart of a data processing method for predicting tunneling parameters provided in this application;
[0048] Figure 3 A flow chart of a data processing method for predicting tunneling parameters provided in this application;
[0049] Figure 4 A flow chart of a data processing method for predicting tunneling parameters provided in this application;
[0050] Figure 5 A schematic diagram of a data processing device for predicting tunneling parameters provided in the present application. DETAILED DESCRIPTION
[0051] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.
[0052] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0053] In the present application, the terms "upper", "lower", "left", "right", "front", "back", "top", "bottom", "inner", "outer", "middle", "vertical", "horizontal", "lateral", "longitudinal" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings. These terms are mainly used to better describe the present application and its embodiments, and are not used to limit the indicated devices, elements or components to have a specific orientation, or to be constructed and operated in a specific orientation.
[0054] In addition, some of the above terms may be used to express other meanings in addition to indicating orientation or positional relationship. For example, the term "on" may also be used to express a certain dependency or connection relationship in some cases. For those of ordinary skill in the art, the specific meanings of these terms in this application can be understood according to specific circumstances.
[0055] In addition, the terms "installed", "set", "provided with", "connected", "connected", and "socketed" should be understood in a broad sense. For example, "connection" can be a fixed connection, a detachable connection, or an integral structure; it can be a mechanical connection or an electrical connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be an internal connection between two devices, elements, or components. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.
[0056] In an optional embodiment of the present application, a data processing method for predicting tunneling parameters is proposed. Figure 1 A flow chart of a data processing method for predicting tunneling parameters provided in this application, such as Figure 1 As shown, the method comprises the following steps:
[0057] S101: Obtain data to be processed;
[0058] The data to be processed are related data used to represent tunneling construction parameters. The tunneling construction parameters include raw data from the tunneling process, including data continuously collected by various sensors installed on the TBM at a uniform time interval, such as 1s.
[0059] S102: Performing feature parameter screening processing on the data to be processed based on causal discovery to obtain input feature parameter data;
[0060] The input characteristic parameter data is data for representing characteristic parameters for predicting tunneling parameters in the data to be processed. The data to be processed is subjected to global causal discovery processing to achieve feature screening and filtering in the data to be processed, and the global causal relationship and connection strength between the various parameters of the TBM are obtained to realize the input data of the tunneling parameter prediction model in the data to be processed, thereby realizing the prediction processing of the parameters to be predicted. In the embodiment of the present application, the input characteristic parameters whose causal relationship strength meets the causal relationship strength rule are screened through global causal discovery, so as to improve the accuracy and prediction precision of tunneling parameter prediction.
[0061] In an optional embodiment of the present application, a data processing method for predicting tunneling parameters is proposed. Figure 2 A flow chart of a data processing method for predicting tunneling parameters provided in this application, such as Figure 2 As shown, the method comprises the following steps:
[0062] S201: performing normalization-based data preprocessing on the data to be processed to obtain preprocessed tunneling parameter data;
[0063] The pre-processed tunneling parameter data is used to represent the data to be processed after normalization. The normalization-based data pre-processing of the data to be processed includes: screening the abnormal data to be processed, cleaning the downtime data and abnormal data, etc.; normalizing the data to be processed after cleaning the abnormal data, the formula is:
[0064]
[0065] Among them, pi is the TBM parameter before normalization, pi* is the normalized TBM parameter, and the TBM parameters after filtering abnormal data are normalized by the above formula to obtain the above-mentioned preprocessing excavation parameter data.
[0066] S202: performing causal relationship strength analysis processing based on causal discovery on the pre-processed tunneling parameter data to obtain parameter causal relationship strength characteristic data;
[0067] The parameter causal relationship strength feature data is feature data used to represent the causal relationship strength of feature parameters.
[0068] In an optional embodiment of the present application, causal discovery is performed on the pre-processed tunneling parameter data, including: calculating residuals, independence, determining exogenous variables, and calculating connection strength on the pre-processed tunneling parameter data in sequence, as follows:
[0069] First, the mean is subtracted from each row of the data matrix X, and then the Ica algorithm is applied to obtain the decomposition X = AS, where S has the same size as X and contains independent components in its rows. The mixing matrix W = A-1 is estimated;
[0070] Find the unique permutation of the rows of W that produces a matrix without any zeros on the main diagonal In practice, small estimation errors will cause all elements of W to be nonzero, so we seek to make Minimized permutations;
[0071] Will Divide each row of by its corresponding diagonal element to produce a new matrix with all 1s on the diagonal
[0072] Calculate an estimate of B
[0073]
[0074] Find the causal order, that is, find The permutation matrix P of (applying the same to rows and columns) produces the matrix As close as possible to a strict lower triangle, use To measure;
[0075] Analyze the cause-effect relationship between TBM parameters, HS t =HS t-1 , calculate the causal relationship strength of each parameter respectively, and obtain the causal relationship strength characteristic data of the above parameters.
[0076] S203: Screening the parameter causal relationship strength feature data based on a preset causal relationship strength threshold to obtain input feature parameter data.
[0077] In an optional embodiment of the present application, a data processing method for predicting excavation parameters is proposed, the method comprising: comparing parameter causal relationship strength characteristic data with a preset causal relationship strength threshold to implement screening processing of the characteristic parameters, if the strength value of the parameter causal relationship strength characteristic data is greater than the preset causal relationship strength threshold, obtaining input characteristic data; matching parameter data corresponding to the input characteristic data in the pre-processed excavation parameters to obtain input characteristic parameter data.
[0078] In an embodiment of the present application, according to the relationship strength network, feature quantities whose absolute values of relationship strength are greater than a preset causal relationship strength threshold are screened as input features. For example, different feature parameters correspond to different causal relationship strengths, and the preset causal relationship strength threshold is the absolute value of the relationship strength of 0.1. When the absolute value of the causal relationship strength of the feature parameter is greater than 0.1 and not equal to 1, the feature parameter is screened as the input feature, and all data corresponding to the feature parameter is obtained in the data to be processed as input data for the tunneling parameter prediction model.
[0079] S103: Perform parameter prediction processing on the input characteristic parameter data based on the tunneling parameter prediction model to obtain predicted tunneling parameter data.
[0080] In an optional embodiment of the present application, a data processing method for predicting tunneling parameters is proposed. Figure 3 A flow chart of a data processing method for predicting tunneling parameters provided in this application, such as Figure 3 As shown, the method comprises the following steps:
[0081] S301: performing causal feature extraction processing based on causal discovery on input feature parameter data to obtain input causal feature data;
[0082] The input causal feature data is feature data used to represent the causal relationship in the input feature parameters.
[0083] In an embodiment of the present application, causal discovery processing is performed on the input feature parameter data, and based on the same causal discovery steps as mentioned above, local causal discovery is performed on the input feature parameter data to obtain an adjacency matrix, which is the input causal feature data.
[0084] Performing local causal discovery on the input feature parameter data to obtain an adjacency matrix includes: performing cyclic segment segmentation processing on the above preprocessed data to obtain multiple cyclic segment data, performing local causal discovery processing on each cyclic segment data to obtain multiple local adjacency matrices; splicing the multiple local adjacency matrices into the corresponding cyclic segment data. For example, if there are m data, each data includes n feature quantities, so that the original data is an m×n matrix, and after performing causal discovery on the above m×n matrix, an n×n local adjacency matrix will be obtained, and the two matrices are spliced to obtain an m×n×n matrix to obtain the above adjacency matrix.
[0085] S302: performing data splicing processing on the input causal feature data and the input feature parameter data to obtain spliced input parameter data;
[0086] The above-mentioned adjacency matrix and the above-mentioned input characteristic parameter data are connected as input data of the tunneling parameter prediction model to realize the prediction processing of the tunneling parameters.
[0087] S303: Perform parameter prediction processing on the spliced data parameter data based on the tunneling parameter prediction model to obtain predicted tunneling parameter data.
[0088] In an optional embodiment of the present application, by performing global causal discovery on the data to be processed, the screening processing of the input feature parameters is achieved, and by performing local causal discovery on the input feature parameter data, the causal relationship mining of the feature parameters is achieved, and a local adjacency matrix is obtained. The local adjacency matrix is used as a new model input and together with the above-mentioned input feature parameters are used as the input of the excavation parameter prediction model to achieve target excavation parameter prediction, thereby improving the accuracy and precision of the model prediction.
[0089] In an optional embodiment of the present application, a data processing method for predicting tunneling parameters is proposed. Figure 4 A flow chart of a data processing method for predicting tunneling parameters provided in this application, such as Figure 4 As shown, the method comprises the following steps:
[0090] S401: Obtain sample data to be processed;
[0091] The sample data to be processed are excavation parameter sample data used to train the excavation parameter prediction model. The sample data are original data obtained from the tunnel excavation process, including sample data continuously collected at uniform time intervals by various sensors installed on the TBM.
[0092] S402: Performing feature parameter screening processing on the sample data to be processed based on causal discovery to obtain sample input feature parameter data;
[0093] The sample input feature parameter data is data used to represent the input feature parameters of the model to be trained in the sample data to be processed.
[0094] In an optional embodiment of the present application, the sample data to be processed is normalized, including normalizing the sample data after screening based on abnormal data; global causal discovery is performed on the normalized sample data based on the above-mentioned causal discovery method to implement screening of sample input feature parameters, screen input feature parameters that meet the parameter causal relationship strength, and obtain data of the input feature parameters in the sample data as input data for training the tunneling parameter prediction model.
[0095] S403: Perform model training processing on a preset deep learning model according to the sample input characteristic parameter data and the sample data to be processed to obtain a tunneling parameter prediction model.
[0096] In an optional embodiment of the present application, a data processing method for predicting tunneling parameters is proposed, comprising:
[0097] The sample input feature parameter data is subjected to causal feature extraction processing based on causal discovery to obtain sample input causal feature data, wherein the sample input causal feature data is feature data used to represent the causal relationship in the sample input feature parameters; the sample input causal feature data and the sample input feature parameter data are subjected to data splicing processing to obtain sample splicing input parameter data; the sample data to be processed is subjected to output feature parameter extraction processing to obtain sample output feature parameter data; the sample splicing input parameter data is used as the input sample output feature parameter data of a preset deep learning model as the output of the preset deep learning model, and the preset deep learning model is subjected to model training processing to obtain a tunneling parameter prediction model.
[0098] The excavation parameter prediction model is trained through training sample data, and data feature parameter extraction and processing are performed on the sample data to be processed. The input features are set based on the prediction requirements. For example, during tunnel excavation construction, it is necessary to predict construction parameters A and construction parameters B. The construction parameters A and construction parameters B are used as the output data of the excavation parameter prediction model. The data corresponding to construction parameters A and construction parameters B in the samples to be processed are identified as the output of the model to be trained. According to the model input data and output data, the preset deep learning model is trained to obtain the excavation parameter prediction model for predicting construction parameters A and construction parameters B.
[0099] In an optional embodiment of the present application, the structure of the preset deep learning model consists of an embedding layer, multiple transformer layers and an output layer. One of the Transformer Blocks contains a self-attention layer, layer normalization, a feedforward network, a residual connection and Dropout. The embedding layer maps the input features to the embedding space, and the linear layer maps the input data from the original feature dimension to the embedding dimension. Multiple Transformer Blocks are stacked to capture more complex features and contextual relationships. The output layer maps the output of the Transformer Block to the predicted dimension to obtain real-time prediction results of the TBM parameters.
[0100] In an optional embodiment of the present application, a data processing method for predicting tunneling parameters is proposed, comprising:
[0101] The sample splicing input parameter data is used as the input of the preset deep learning model, and the sample output feature parameter data is used as the output of the preset deep learning model. The preset deep learning model is subjected to model training processing to obtain a process excavation parameter prediction model; the process excavation parameter prediction model is subjected to model evaluation processing to obtain model evaluation feature data, wherein the model evaluation feature data is feature data used to represent the performance of the process excavation parameter prediction model; the model evaluation feature data is subjected to judgment processing based on preset performance evaluation rules, and if the model evaluation feature data satisfies the preset performance evaluation rules, an excavation parameter prediction model is obtained, wherein the excavation parameter prediction model is a process excavation parameter prediction model that satisfies the preset performance evaluation rules; if the model evaluation feature data does not satisfy the preset performance evaluation rules, the process excavation parameter prediction model is subjected to model iteration training processing until the process excavation parameter prediction model satisfies the preset performance evaluation rules to obtain the excavation parameter prediction model.
[0102] The process excavation parameter prediction model was evaluated by testing sample data, and the mean square error (MSE), mean absolute error (MAE) and R square score (R 2 score), and train the model based on MSE as the loss function.
[0103]
[0104]
[0105]
[0106] Among them, y i represents the output of the tunneling parameter prediction model for the above process, Represents y i The corresponding value in the test data set, Y represents the mean of the output results.
[0107] In an optional embodiment of the present application, a data processing method for predicting tunneling parameters is proposed, comprising:
[0108] The excavation parameter prediction model obtained through the above training is subjected to model interpretation processing, including:
[0109] The excavation parameter prediction model is calculated based on the sampling approximate method SHapley AdditiveexPlanations (hereinafter referred to as SHAP value). The specific calculation process is as follows:
[0110]
[0111] Among them, SV irepresents the contribution of a single eigenvalue, F represents the set of features in the dataset, x represents a sample of the dataset, Π(N) represents the set of all ordered permutations of N, and p(O) represents the set of predecessors of player i, whose order is Q∈(N).
[0112] In an embodiment of the present application, by performing model interpretation processing on the excavation parameter prediction model, the influence of each feature of each time step and single sample in the model on the final prediction result is realized, so as to optimize the model during the training process of the excavation parameter prediction model; through the SHAP value, the features that have an important influence on the model prediction results are identified, so as to perform feature selection.
[0113] In an optional embodiment of the present application, a sampling-based approximation method is used to calculate the SHAP value, explore the impact of the selected features on the prediction results, and quantify the extent of the impact, thereby achieving the interpretability of the model, facilitating parameter adjustment and decision-making in tunneling construction, and improving the stability and safety of the tunneling machine excavation process.
[0114] In an optional embodiment of the present application, a data processing device for predicting tunneling parameters is proposed. Figure 5 A schematic diagram of a data processing device for predicting tunneling parameters provided in this application, such as Figure 5 Shown include:
[0115] A data acquisition module 51 is used to acquire data to be processed, wherein the data to be processed is relevant data used to represent excavation construction parameters;
[0116] The characteristic parameter screening module 52 is used to perform characteristic parameter screening processing based on causal discovery on the data to be processed to obtain input characteristic parameter data, wherein the input characteristic parameter data is data for representing characteristic parameters for predicting tunneling parameters in the data to be processed;
[0117] The excavation parameter prediction module 53 is used to perform parameter prediction processing on the input characteristic parameter data based on the excavation parameter prediction model to obtain predicted excavation parameter data.
[0118] The specific manner of executing the operation of each unit in the above embodiment has been described in detail in the embodiment of the method, and will not be elaborated here.
[0119] In summary, the data to be processed is obtained, and the data to be processed is the relevant data used to represent the excavation construction parameters; the data to be processed is subjected to characteristic parameter screening processing based on causal discovery to obtain input characteristic parameter data, wherein the input characteristic parameter data is the data used to represent characteristic parameters for predicting excavation parameters in the data to be processed; the input characteristic parameter data is subjected to parameter prediction processing based on the excavation parameter prediction model to obtain predicted excavation parameter data. By performing causal analysis on the excavation construction parameters, the excavation parameter prediction model performs parameter prediction processing on the causal relationship characteristics and the relevant excavation parameter data, thereby solving the problem of low accuracy of TBM excavation parameter prediction in the prior art and achieving the technical effect of improving the accuracy of excavation parameter prediction.
[0120] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.
[0121] Obviously, those skilled in the art should understand that the above-mentioned units or steps of the present application can be implemented by a general-purpose computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. In this way, the present application is not limited to any specific combination of hardware and software.
[0122] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims
1. A data processing method for predicting tunneling parameters, characterized in that: include: Acquiring data to be processed, wherein the data to be processed is relevant data for representing excavation construction parameters; Performing feature parameter screening processing based on causal discovery on the data to be processed to obtain input feature parameter data, wherein the input feature parameter data is data for representing feature parameters for predicting tunneling parameters in the data to be processed; The input characteristic parameter data is subjected to parameter prediction processing based on a tunneling parameter prediction model to obtain predicted tunneling parameter data.
2. The data processing method according to claim 1, characterized in that: The data to be processed is subjected to feature parameter screening based on causal discovery, and the input feature parameter data obtained includes: Performing normalized data preprocessing on the data to be processed to obtain preprocessed tunneling parameter data, wherein the preprocessed tunneling parameter data is used to represent the data to be processed after normalization processing; Performing causal relationship strength analysis processing based on causal discovery on the pre-processed excavation parameter data to obtain parameter causal relationship strength characteristic data, wherein the parameter causal relationship strength characteristic data is characteristic data used to represent the causal relationship strength of characteristic parameters; The parameter causal relationship strength characteristic data is screened based on a preset causal relationship strength threshold to obtain the input characteristic parameter data.
3. The data processing method according to claim 2, characterized in that: The parameter causal relationship strength characteristic data is screened based on a preset causal relationship strength threshold to obtain the input characteristic parameter data including: Comparing the parameter causal relationship strength characteristic data with the preset causal relationship strength threshold to implement screening processing of characteristic parameters, If the strength value of the parameter causal relationship strength feature data is greater than the preset causal relationship strength threshold, obtaining input feature data; Parameter data corresponding to the input characteristic data are matched in the pre-processing excavation parameters to obtain the input characteristic parameter data.
4. The data processing method according to claim 1, characterized in that: The input characteristic parameter data is subjected to parameter prediction processing based on the excavation parameter prediction model to obtain predicted excavation parameter data including: Performing causal feature extraction processing based on causal discovery on the input feature parameter data to obtain input causal feature data, wherein the input causal feature data is feature data used to represent the causal relationship in the input feature parameters; Performing data splicing processing on the input causal feature data and the input feature parameter data to obtain spliced input parameter data; The spliced data parameter data is subjected to parameter prediction processing based on a tunneling parameter prediction model to obtain the predicted tunneling parameter data.
5. The data processing method according to claim 1, characterized in that: Before obtaining the data to be processed, the method further includes: Acquire sample data to be processed, wherein the sample data to be processed is excavation parameter sample data used to train the excavation parameter prediction model; Performing feature parameter screening processing based on causal discovery on the sample data to be processed to obtain sample input feature parameter data, wherein the sample input feature parameter data is data used to represent input feature parameters of the model to be trained in the sample data to be processed; The preset deep learning model is trained according to the sample input characteristic parameter data and the sample data to be processed to obtain the tunneling parameter prediction model.
6. The data processing method according to claim 5, characterized in that: The preset deep learning model is trained according to the sample input characteristic parameter data and the sample data to be processed to obtain the tunneling parameter prediction model, which includes: Performing causal feature extraction processing based on causal discovery on the sample input feature parameter data to obtain sample input causal feature data, wherein the sample input causal feature data is feature data used to represent the causal relationship in the sample input feature parameters; Performing data splicing processing on the sample input causal feature data and the sample input feature parameter data to obtain sample splicing input parameter data; Performing output characteristic parameter extraction processing on the sample data to be processed to obtain sample output characteristic parameter data; The sample splicing input parameter data is used as the input of the preset deep learning model, and the sample output feature parameter data is used as the output of the preset deep learning model. The preset deep learning model is subjected to model training processing to obtain the excavation parameter prediction model.
7. The data processing method according to claim 6, characterized in that: The sample splicing input parameter data is used as the input of the preset deep learning model, the sample output characteristic parameter data is used as the output of the preset deep learning model, and the preset deep learning model is trained to obtain the tunneling parameter prediction model, which includes: The sample splicing input parameter data is used as the input of the preset deep learning model, and the sample output characteristic parameter data is used as the output of the preset deep learning model, and the preset deep learning model is subjected to model training processing to obtain a process excavation parameter prediction model; Performing model evaluation processing on the process excavation parameter prediction model to obtain model evaluation characteristic data, wherein the model evaluation characteristic data is characteristic data used to represent the performance of the process excavation parameter prediction model; The model evaluation feature data is judged based on preset performance evaluation rules. If the model evaluation characteristic data satisfies the preset performance evaluation rules, the excavation parameter prediction model is obtained, wherein the excavation parameter prediction model is a process excavation parameter prediction model that satisfies the preset performance evaluation rules; If the model evaluation feature data does not satisfy the preset performance evaluation rules, the process excavation parameter prediction model is subjected to model iteration training processing until the process excavation parameter prediction model satisfies the preset performance evaluation rules, thereby obtaining the excavation parameter prediction model.
8. A data processing device for predicting tunneling parameters, characterized in that: include: A data acquisition module, used to acquire data to be processed, wherein the data to be processed is relevant data used to represent excavation construction parameters; A characteristic parameter screening module is used to perform characteristic parameter screening processing on the data to be processed based on causal discovery to obtain input characteristic parameter data, wherein the input characteristic parameter data is data for representing characteristic parameters for predicting tunneling parameters in the data to be processed; The excavation parameter prediction module is used to perform parameter prediction processing on the input characteristic parameter data based on the excavation parameter prediction model to obtain predicted excavation parameter data.
9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the data processing method for tunneling parameter prediction according to any one of claims 1 to 7.
10. An electronic device, characterized in that: include: at least one processor; And a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor executes the data processing method for excavation parameter prediction as described in any one of claims 1-7.