CN115422821A - Data processing method and device for rock mass parameter prediction - Google Patents

Abstract

The application discloses a data processing method and device for rock mass parameter prediction. The method includes: acquiring data to be processed; performing preprocessing based on data analysis on the data to be processed to obtain excavation characteristic data; performing rock mass parameter prediction processing on the excavation characteristic data based on a preset rock mass parameter prediction model, Obtain the parameter data of the target rock mass. The excavation feature data is obtained by preprocessing the obtained data to be processed based on data analysis; and the rock mass parameter prediction processing is performed on the excavation feature data based on a preset rock mass parameter prediction model to obtain target rock mass parameter data. The rock mass parameter prediction of the excavation characteristic data is carried out through the preset rock mass parameter model, which solves the technical problem of low accuracy of rock mass parameter judgment in the process of TBM tunnel construction in the prior art, and realizes the improvement of rock mass parameter prediction The technical effect of accuracy.

Description

Data processing method and device for rock mass parameter prediction

technical field

The present application relates to the field of computers, in particular, to a data processing method and device for rock mass parameter prediction.

Background technique

The key to the realization of safe and efficient excavation of full-face hard rock tunnel boring machine (Tunnel Boring Machine, TBM for short) lies in the real-time and accurate perception of the rock mass characteristics of the tunnel face during the excavation process of TBM. The most reasonable TBM excavation strategy can be adopted only after the on-site construction personnel have an accurate judgment on the condition of the rock mass at the face, so as to realize safe and efficient TBM excavation. However, the TBM cutterhead and the shield almost cut off all the rock mass information within 2m of the tunnel face and the surrounding cave walls. Although part of the surrounding rock of the tunnel face can be directly observed through the gap of the hob support of the manual tool change chamber, During the excavation process, the tool change chamber often has splashed slag pieces, and the slag ash seriously reduces the visibility inside the tool change chamber, and the judgment efficiency of rock mass parameters is low, and the judgment of rock mass parameters is not accurate.

Therefore, in the prior art, there is a problem of low accuracy in judging rock mass parameters during TBM tunnel construction.

Contents of the invention

The main purpose of this application is to provide a data processing method and device for predicting rock mass parameters, so as to solve the technical problem of low accuracy in judging rock mass parameters during the construction of TBM tunnels in the prior art, and improve rock mass parameters. The accuracy of rock mass parameter prediction is convenient for adjusting the TBM tunnel construction strategy according to the rock mass parameters, thereby improving the efficiency of TBM tunnel construction.

In order to achieve the above object, the first aspect of the present application proposes a data processing method for rock mass parameter prediction, including:

Obtaining the data to be processed, wherein the data to be processed is relevant data used to represent the excavation construction of the excavation equipment in the rock mass to be predicted;

Preprocessing the data to be processed based on data analysis to obtain excavation characteristic data, wherein the excavation characteristic data is data representing the excavation characteristics of the excavation equipment; and

Based on a preset rock mass parameter prediction model, rock mass parameter prediction processing is performed on the excavation feature data to obtain target rock mass parameter data.

Optionally, performing rock mass parameter prediction processing on the excavation feature data based on a preset rock mass parameter prediction model, to obtain target rock mass parameter data including:

Perform a first prediction process on the excavation feature data based on a preset first rock mass parameter model to obtain first rock mass parameter data, wherein the first rock mass parameter data is used to represent the rock mass to be predicted data of the first process prediction parameter;

performing characteristic analysis and processing on the excavation characteristic data and the first rock mass parameter data to obtain process characteristic data;

The second prediction process is performed on the process feature data based on the second preset rock mass parameter model to obtain second rock mass parameter data, wherein the second rock mass parameter data is used to represent the rock mass parameter to be predicted the data of the second process prediction parameter; and

Perform verification processing on the second rock mass parameter data to obtain the target rock mass parameter data.

Optionally, performing feature analysis and processing on the excavation feature data and the first rock mass parameter data to obtain process feature data includes:

Identifying the rock mass integrity feature of the first rock mass parameter to obtain rock mass integrity feature data, wherein the rock mass integrity feature data is data representing the integrity feature of the rock mass to be predicted;

Perform rock mass integrity classification and prediction processing on the rock mass integrity feature data to obtain classification feature data, wherein the classification feature data includes rock mass integrity categories and corresponding category probabilities; and

The classification feature data is matched with the excavation feature data to obtain the process feature data.

Optionally, preprocessing the data to be processed based on data analysis to obtain the excavation characteristic data includes:

Obtaining reference data, wherein the reference data is data related to the excavation of the excavation equipment in an unloaded state;

Filtering the data to be processed based on the reference data to obtain process data to be processed, wherein the process data to be processed is the data to be processed after filtering the reference data; and

Based on preset feature extraction rules, feature extraction processing is performed on the data to be processed in the process to obtain the excavation feature data, wherein the preset feature extraction rules correspond to excavation features, and the excavation feature data are in the process The data of excavation features extracted from the data to be processed.

Optionally, before obtaining the data to be processed, the method further includes:

Acquiring training sample data, wherein the training sample data is sample data for training the preset rock mass parameter prediction model;

Preprocessing the training sample data based on data analysis to obtain a plurality of training characteristic data, wherein the plurality of training characteristic data are characteristic data associated with rock mass parameters; and

The preset neural network model is trained based on the plurality of training feature data to obtain the preset rock mass parameter prediction model.

Optionally, the preset neural network model is trained based on the plurality of training feature data, and the preset rock mass parameter prediction model obtained includes:

performing classification processing on the plurality of training feature data to obtain a first training feature data set and a second training feature data set, wherein the first training feature data set includes a plurality of first training feature data, and the second The training feature data set includes a plurality of second training feature data;

According to the first training feature data set, the preset regression classification model is trained to obtain the first rock mass prediction model, wherein the rock mass classification prediction model is used to classify rock mass integrity Model;

Perform rock mass integrity feature prediction processing on the second training feature data set based on the first rock mass prediction model to obtain process training feature data, wherein the process training feature data is used to represent rock mass integrity characteristic data; and

The preset neural network model is trained according to the second training feature data set and the process training feature data to obtain the rock mass parameter prediction model.

According to the second aspect of the present application, a data processing device for rock mass parameter prediction is proposed, including:

The data acquisition module is used to acquire data to be processed, wherein the data to be processed is relevant data used to represent the excavation construction of the excavation equipment in the rock mass to be predicted;

A preprocessing module, configured to preprocess the data to be processed based on data analysis to obtain excavation characteristic data, wherein the excavation characteristic data is data representing the excavation characteristics of the excavation equipment; and

The prediction module performs rock mass parameter prediction processing on the excavation feature data based on a preset rock mass parameter prediction model to obtain target rock mass parameter data.

Optionally, the prediction module includes:

The first prediction module performs a first prediction process on the excavation feature data based on a preset first rock mass parameter model to obtain first rock mass parameter data, wherein the first rock mass parameter data is used to represent the The data of the first process prediction parameter of the rock mass to be predicted;

performing characteristic analysis and processing on the excavation characteristic data and the first rock mass parameter data to obtain process characteristic data;

The second prediction module performs a second prediction process on the process feature data based on a preset second rock mass parameter model to obtain second rock mass parameter data, wherein the second rock mass parameter data is used to represent the data of the second process prediction parameter of the rock mass parameter to be predicted; and

The result module is configured to perform verification processing on the second rock mass parameter data to obtain the target rock mass parameter data.

According to the third aspect of the present application, a computer-readable storage medium is proposed, the computer-readable storage medium stores computer instructions, and the computer instructions are used to make the computer execute the above-mentioned method for rock mass parameter prediction data processing method.

According to a fourth aspect of the present application, an electronic device is proposed, including: at least one processor; and a memory connected to the at least one processor in communication; wherein, the memory stores information that can be used by the at least one processor Executable computer program, the computer program is executed by the at least one processor, so that the at least one processor executes the above data processing method for rock mass parameter prediction.

The technical solutions provided by the embodiments of the present application may include the following beneficial effects:

In this application, the excavation characteristic data is obtained by preprocessing the acquired data to be processed based on data analysis; the rock mass parameter prediction processing is performed on the excavation characteristic data based on the preset rock mass parameter prediction model, and the target rock mass is obtained. Body parameter data. Through the processing and filtering of the excavation characteristic data, combined with the rock mass parameter prediction model, the rock mass parameter data with higher accuracy is predicted, which solves the problem of low accuracy of rock mass parameter judgment in the TBM tunnel construction process in the prior art. Technical problems have been solved, and the technical effect of improving the accuracy of rock mass parameter prediction has been achieved.

Description of drawings

The accompanying drawings, which constitute a part of this application, are included to provide a further understanding of the application and make other features, objects and advantages of the application apparent. The drawings and descriptions of the schematic embodiments of the application are used to explain the application, and do not constitute an improper limitation to the application. In the attached picture:

Fig. 1 is a schematic flow chart of a data processing method for rock mass parameter prediction provided by the present application;

Fig. 2 is a schematic flow chart of a data processing method for rock mass parameter prediction provided by the application;

Fig. 3 is a schematic flow chart of a data processing method for rock mass parameter prediction provided by the present application;

Fig. 4 is a schematic flow chart of a data processing method for rock mass parameter prediction provided by the present application;

Fig. 5 is a schematic structural diagram of a data processing device for rock mass parameter prediction provided by the present application;

FIG. 6 is a schematic structural diagram of another data processing device for rock mass parameter prediction provided by the present application.

Detailed ways

In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiment of the application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiment of the application. Obviously, the described embodiment is only It is an embodiment of a part of the application, but not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without creative efforts shall fall within the scope of protection of this application.

It should be noted that the terms "first" and "second" in the description and claims of the present application and the above drawings are used to distinguish similar objects, but not necessarily used to describe a specific sequence or sequence. It should be understood that the data so used may be interchanged under appropriate circumstances for the embodiments of the application described herein. Furthermore, the terms "comprising" and "having", as well as any variations thereof, are intended to cover a non-exclusive inclusion, for example, a process, method, system, product or device comprising a sequence of steps or elements is not necessarily limited to the expressly listed instead, may include other steps or elements not explicitly listed or inherent to the process, method, product or apparatus.

In this application, the terms "upper", "lower", "left", "right", "front", "rear", "top", "bottom", "inner", "outer", "middle", The orientations or positional relationships indicated by "vertical", "horizontal", "horizontal", and "longitudinal" are based on the orientations 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 that the indicated devices, elements or components must have a specific orientation, or be constructed and operated in a specific orientation.

Moreover, some of the above terms may be used to indicate other meanings besides orientation or positional relationship, for example, the term "upper" may also be used to indicate a certain attachment relationship or connection relationship in some cases. Those skilled in the art can understand the specific meanings of these terms in this application according to specific situations.

Furthermore, the terms "installed", "disposed", "provided", "connected", "connected", "socketed" are to be interpreted broadly. 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 intermediary, or two devices, components or Internal connectivity between components. Those of ordinary skill in the art can understand the specific meanings of the above terms in this application according to specific situations.

Fig. 1 provides a kind of data processing method for rock mass parameter prediction provided by the application, as shown in Fig. 1, this method comprises the following steps:

S101: Obtain data to be processed;

The data to be processed is relevant data used to indicate that the excavation equipment is excavating the rock mass to be predicted. During the excavation construction of the rock mass, the excavation equipment generates various types of data, including the first construction data and the second construction data. The evaluation data to be processed can be divided into active construction data and passive construction data according to the form of data generation. The active construction data is used to represent the parameter data used to control the construction of the tunneling equipment, and the passive construction data is used to represent the tunneling equipment during the construction process. The data generated by the construction impact, such as active construction data including the preset construction parameters of the tunneling equipment such as cutterhead speed, cutterhead torque, construction power, etc., passive construction data is obtained by collecting the vibration characteristics of the tunneling equipment during the construction process , such as the collected vibration data of the cutter head. When collecting excavation equipment, the vibration data of the cutter head is collected through the TBM vibration monitoring system installed on the excavation equipment, and the sensor module in the TBM vibration monitoring system is used to collect and store the data in real time, and the sensor module collects the data in real time The stored data is passed to the server.

S102: Preprocessing the data to be processed based on data analysis to obtain excavation feature data;

The excavation feature data is the data representing the excavation characteristics of the excavation equipment. The obtained data to be processed is the relevant data obtained during the preset time period of the current predicted rock mass construction. The corresponding excavation equipment excavation at any time within the preset time period The construction data is difficult to intuitively reflect the excavation characteristics in the current time period. It is necessary to process the data generated by the excavation equipment excavation construction in a certain period of time based on data analysis to obtain the excavation characteristic data, which is convenient for the current rock mass to be predicted. The tunneling characteristics of the tunneling construction process are directly reflected, the data processing efficiency of the rock mass parameter prediction process is improved, and the accuracy of the rock mass parameter prediction by the rock mass parameter model is improved.

Fig. 2 is a kind of data processing method for rock mass parameter prediction that the present application provides, as shown in Fig. 3, this method comprises the following steps:

S201: Obtain reference data;

The reference data is the data related to the excavation of the excavation equipment under no-load condition. For example, the reference data includes the host frequency of the excavation equipment. Vibration frequency generated during non-drilling construction.

S202: Filter the data to be processed based on the reference data to obtain the data to be processed in the process;

Filter the reference data in the data to be processed. The data to be processed includes the construction parameter data to be processed and the excavation vibration data to be processed. The excavation vibration data to be processed is filtered through the main engine frequency in the reference data, and the excavation construction with the main engine frequency filtered out is obtained. The vibration data of the cutter head will filter out the data to be processed from the reference data as the data to be processed in the process.

S203: Perform feature extraction processing on the data to be processed in the process based on preset feature extraction rules to obtain excavation feature data;

峰值X_p，X_p＝max{|x_i|}，振动加速度有效值X_RMS，

平均极值，为该岩体状态下预设个数的振动极大值和振动极小值计算绝对值平均值，其中，x_i为预设时间段内的第i个刀盘振动数据，T为预设时间段内的时长，N为TBM在预设时间段内共采集的振动数据的个数，上述刀盘振动的平均幅值，峰值，振动加速度有效值和平均极值为掘进特征数据中的掘进振动特征；对待处理参数中的施工特征数据进行特征提取，得到刀盘转速、刀盘扭矩、贯入度和切深指数FPI，为掘进特征数据中的掘进施工特征数据。通过对待处理数据进行特征提取的数据预处理，提高了进行岩体参数预测模型的数据数据规范性和准确率，进而实现提高对目标岩体参数数据预测的准确性。The preset feature extraction rules correspond to the excavation features. The excavation feature data is the data of the excavation features extracted from the data to be processed in the process. The excavation feature parameters include the excavation vibration characteristic parameters and the excavation construction characteristic parameters. For the filtered TBM Carry out data analysis on the vibration data of the cutterhead, count the vibration data of the TBM cutterhead within the preset time period, and obtain the average amplitude of the cutterhead vibration

Peak value X _p , X _p ＝max{| _xi |}, effective value of vibration acceleration X _RMS ,

The average extreme value is to calculate the absolute value average value for the preset number of vibration maximum values and vibration minimum values in the state of the rock mass, where x _i is the i-th cutterhead vibration data within the preset time period, and T is the length of time in the preset time period, N is the number of vibration data collected by the TBM in the preset time period, the average amplitude, peak value, effective value and average extreme value of the vibration acceleration of the above-mentioned cutterhead are the excavation characteristic data The excavation vibration characteristics in the excavation feature data; the construction characteristic data in the parameters to be processed are extracted to obtain the cutterhead speed, cutterhead torque, penetration and depth of cut index FPI, which are the excavation construction characteristic data in the excavation characteristic data. Through the data preprocessing of the feature extraction of the data to be processed, the data standardization and accuracy of the rock mass parameter prediction model are improved, and the accuracy of the target rock mass parameter data prediction is improved.

S103: Perform rock mass parameter prediction processing on the excavation feature data based on the preset rock mass parameter prediction model to obtain target rock mass parameter data;

Fig. 3 is a kind of data processing method for rock mass parameter prediction that the present application provides, as shown in Fig. 2, this method comprises the following steps:

S301: Perform first prediction processing on the excavation characteristic data based on the preset first rock mass parameter model to obtain first rock mass parameter data;

The first rock mass parameter data is used to represent the data of the first process prediction parameter of the rock mass to be predicted;

The rock mass integrity feature of the first rock mass parameter is identified to obtain the rock mass integrity feature data, the rock mass integrity feature data is the data used to represent the integrity feature of the rock mass to be predicted; the rock mass integrity feature data Perform rock mass integrity classification and prediction processing to obtain classification feature data, which includes rock mass integrity categories and corresponding category probabilities;

The classification characteristic data is matched with the excavation characteristic data to obtain the process characteristic data.

Classify and predict rock mass integrity, and obtain rock mass integrity labels corresponding to rock mass integrity feature data, for example, there are five rock mass integrity labels for rock mass integrity classification, and multiple rock mass integrity Integrity label, by returning the prediction function predict_proba to obtain the probability values of the sample data under multiple rock mass integrity labels, and the category corresponding to the highest probability is the classification label result of the sample. predict_proba returns an array of n rows and k columns. The value on the i-th row and j-th column is the probability that the model predicts that the i-th predicted sample is a certain label. The sum of the probabilities of each row is 1. The probability value corresponding to the label is the largest rock mass integrity label, and the classification feature data is obtained.

S302: Perform characteristic analysis and processing on the excavation characteristic data and the first rock mass parameter data to obtain process characteristic data;

The process feature data is constructed by the excavation feature data and the rock mass integrity classification feature data predicted by the first rock mass parameter model, and the process feature data is obtained according to the corresponding relationship between the excavation feature data and the rock mass integrity classification feature.

S303: Perform a second prediction process on the process feature data based on the preset second rock mass parameter model to obtain second rock mass parameter data;

峰值X_p，平均极值，岩体完整度分类特征数据为经第一岩体参数模型预测的待预测岩体区间内的岩体完整性程度，用于表示待预测岩体的某个区间内的岩体的完整性程度，岩体完整度分类特征数据与掘进特征数据按照岩体掘进区间对应，如第一岩体掘进区间对应的第一岩体完整度分类特征数据，并对应第一掘进特征数据，第二岩体掘进区间对应第二岩体完整度分类特征数据，并对应第二掘进特征数据。The second rock mass parameter model is used to predict the rock mass parameters of the excavation feature data and rock mass integrity classification feature data. The excavation feature data includes excavation construction feature data and excavation vibration feature data. Torque, penetration and depth of cut index FPI, excavation vibration characteristic data include the effective value X _RMS of vibration acceleration in the rock mass interval to be predicted, and the average amplitude

The peak value X _p , the average extreme value, and the rock mass integrity classification characteristic data are the rock mass integrity degree predicted by the first rock mass parameter model in the interval of the rock mass to be predicted, and are used to represent the rock mass integrity in a certain interval of the rock mass to be predicted The integrity degree of the rock mass, the rock mass integrity classification feature data and the excavation feature data correspond to the rock mass excavation interval, for example, the first rock mass integrity classification feature data corresponding to the first rock mass excavation interval corresponds to the first excavation interval The feature data, the second rock mass excavation interval corresponds to the second rock mass integrity classification feature data, and corresponds to the second excavation feature data.

S304: Perform verification processing on the second rock mass parameter data to obtain target rock mass parameter data;

The rock mass parameter data corresponds to a preset rock mass parameter range. By comparing the obtained second rock mass parameter with the preset rock mass parameter range, if the second rock mass parameter is within the preset rock mass parameter range, Judging that the prediction of the second rock mass parameter data is reasonable, the second rock mass parameter is output as the target rock mass parameter; if the second rock mass parameter is not within the preset rock mass parameter range, it is judged that the second rock mass parameter data prediction is unreasonable, The output prompt information of prediction error is convenient to prompt that there is an error in rock mass parameter prediction, so that users can verify the process of rock mass parameter prediction, eliminate problems that affect the verification result, and improve the accuracy of rock mass parameter prediction.

Fig. 4 is a kind of data processing method for rock mass parameter prediction that the present application provides, as shown in Fig. 4, this method comprises the following steps:

S401: Obtain training sample data;

The training sample data is sample data for training a preset rock mass parameter prediction model, and the training sample data includes training tunneling parameter data and training rock mass parameter data.

S402: Preprocessing the training sample data based on data analysis to obtain multiple training feature data;

A plurality of training feature data is a plurality of feature data associated with rock mass parameters, and the training sample data is divided into an interval according to one excavation construction of TBM excavation construction equipment. In this interval, the sample rock mass parameter data of the sample rock mass Corresponding to the sample excavation parameter data, the sample excavation parameter data includes the sample excavation vibration characteristic data and the sample excavation construction characteristic data, the sample excavation vibration characteristic data includes the average amplitude, peak value, vibration acceleration effective value and average extreme value in the sample interval, the sample The characteristic data of tunneling construction include cutterhead speed, cutterhead torque, penetration and depth of cut index FPI in the sample interval; multiple sample interval data sets are constructed based on the sample interval, wherein the multiple sample interval data sets are multiple A training data set of sample excavation construction characteristic data, sample excavation vibration characteristic data and corresponding sample rock mass parameter data.

S403: Perform training on the preset neural network model based on multiple training feature data to obtain a preset rock mass parameter prediction model;

Perform classification processing on a plurality of training feature data to obtain a first training feature data set and a second training feature data set; the first training feature data set includes a plurality of first training feature data, and the second training feature data set includes a plurality of first training feature data sets. Two training feature data; according to the first training feature data set, the preset regression classification model is trained to obtain the first prediction model of rock mass; the rock mass classification prediction model is a model for classifying the integrity of rock mass; based on The first rock mass prediction model performs rock mass integrity feature prediction processing on the second training feature data set to obtain process training feature data; the process training feature data is feature data used to represent rock mass integrity; and according to the second training feature The data set and process training feature data are used to train and process the preset neural network model to obtain a rock mass parameter prediction model.

其中，p₀为准确率，N表示总样本数，N_i表示实际第i类总样本数，

表示预测第i类总样本数。In an optional embodiment of the present application, the rock mass integrity classification model is trained through the first training feature data set. During the training process, the first training feature data set includes sample excavation vibration feature data, sample excavation construction The feature data and the sample rock mass integrity classification feature data, based on the first training feature data set, the preset neural network model is trained and processed, the sample excavation vibration feature data and the sample excavation construction feature data are used as model input, and the sample rock mass The completeness classification feature data is used as the output of the model, and the training of the model parameters is carried out. During the training process, the learning curves of the number of random forest trees, the maximum depth, and the maximum feature are respectively established, the training set is imported, and the first training is carried out. When the learning curve starts to fluctuate, the model enters a stable period; take the first fluctuation range to carry out In the second training, according to the learning curve, the training iteration process is carried out again until the model parameters with higher accuracy are selected, and the target number of random forest trees, the maximum depth of the target and the maximum feature of the target are obtained. Bring the number of target random forest trees, the maximum depth of the target and the maximum feature of the target into the model, perform model evaluation on the model with hyperparameters, obtain the first test sample data, and classify the first test sample data according to the training rock mass integrity classification model Carry out the rock mass integrity classification prediction, evaluate the rock mass integrity classification model through the accuracy rate P and Kappa coefficient k, the accuracy rate P is the proportion of the correct samples predicted by the training rock mass integrity classification model to the first test sample, Kappa The coefficient calculation formula is as follows:

Among them, p ₀ is the accuracy rate, N represents the total number of samples, N _i represents the actual total number of samples of the i-th class,

Indicates the total number of samples predicted for class i.

Performing rock mass integrity feature prediction processing on the second training feature data set based on the first rock mass prediction model to obtain process training feature data;

The first prediction model is used to predict the rock mass integrity feature data of the second training sample to obtain the rock mass integrity feature data of the second training sample, and the predicted rock mass integrity feature data in the second training sample is compared with the first The training excavation construction feature data and training excavation vibration feature data in the second training sample and the sample training rock mass parameters in the second training sample are used as process training feature data, which are feature data for training the rock mass parameter prediction model.

The preset neural network model is trained and processed according to the second training feature data set and the process training feature data to obtain a rock mass parameter prediction model.

The preset neural network model can be a BP neural network model. During the model training process, the hyperparameters in the neural network are trained, and the parameters are sequentially adjusted within the set parameter range by using the grid search method. , use the adjusted parameters to train the model, and find the parameters with the highest accuracy on the test set from the set parameters; by introducing RMSE (Root Mean Squard Error), MAE (Mean Absolute Error) and R ² (RSquared) three An evaluation index is used to evaluate the prediction effect of the model.

表示岩体参数预测值，

表示岩体参数实际值平均值。举例说明，通过采用网格搜索的方式，得到多个超参数对应的多个训练神经网络模型，通过模型评价指标对多个超参数对应的多个训练神经网络模型进行模型评价，确定目标的训练神经网络模型的超参数数据，得到岩体参数预测模型，通过网格搜索后的模型R²值分析比较，针对本数据集学习率为0.3、神经网络的隐藏层层数为1、每隐藏层神经元的个数为3时，神经网络网络效果最佳，根据得到的超参数确定目标训练神经网络模型。Among them, in the formula, y _i represents the actual value of rock mass parameters,

Indicates the predicted value of rock mass parameters,

Indicates the average value of the actual value of the rock mass parameter. For example, by using the grid search method, multiple training neural network models corresponding to multiple hyperparameters are obtained, and the model evaluation index is used to evaluate the multiple training neural network models corresponding to multiple hyperparameters to determine the target training The hyperparameter data of the neural network model is used to obtain the rock mass parameter prediction model. Through the analysis and comparison of the model R ² value after the grid search, the learning rate for this data set is 0.3, the number of hidden layers of the neural network is 1, and each hidden layer When the number of neurons is 3, the effect of the neural network network is the best, and the target training neural network model is determined according to the obtained hyperparameters.

In another optional embodiment of the present application, in training the neural network model, there is determination of the activation function in the neural network model, such as the activation function has Sigmoid (S-type generating curve), Tanh (hyperbolic function), ReLU (linear rectification function), Randomized Leaky ReLU (randomized linear rectification function with leakage correction), etc., to screen training neural network models based on different activation functions. The neural network model with the activation function has the best effect, and the Randomized Leaky ReLU function is selected as the activation function of the neural network model.

In an optional embodiment of the present application, the preset neural network model is trained and processed through the second training feature data set and the process training feature data to obtain a rock mass parameter prediction model, by training the first rock mass prediction model, and by The first rock mass prediction model predicts the integrity classification of the rock mass, and uses the predicted rock mass integrity feature data as the training data for training the rock mass parameter prediction model, by increasing the training data for training the rock mass parameter prediction model The feature dimension improves the accuracy of the trained rock mass parameter model and realizes the accuracy of rock mass parameter prediction.

Fig. 5 is a schematic structural diagram of a data processing device for rock mass parameter prediction provided by the present application. As shown in Fig. 5, the device includes:

The data acquisition module 51 is used to acquire the data to be processed, wherein the data to be processed is relevant data used to indicate that the excavation equipment is excavating the rock mass to be predicted;

A preprocessing module 52, configured to perform preprocessing based on data analysis on the data to be processed to obtain excavation characteristic data, wherein the excavation characteristic data is data used to represent the excavation characteristics of the excavation equipment; and

The prediction module 53 performs rock mass parameter prediction processing on the excavation feature data based on a preset rock mass parameter prediction model, to obtain target rock mass parameter data.

Fig. 6 is a schematic structural diagram of a data processing device for rock mass parameter prediction provided by the present application. As shown in Fig. 6, the device includes:

The first prediction module 61 performs the first prediction processing on the excavation characteristic data based on the preset first rock mass parameter model to obtain the first rock mass parameter data, wherein the first rock mass parameter data is used to represent the rock mass to be predicted data of the first process prediction parameter;

Perform characteristic analysis and processing on the excavation characteristic data and the first rock mass parameter data to obtain process characteristic data;

The second prediction module 62 performs a second prediction process on the process feature data based on the preset second rock mass parameter model to obtain second rock mass parameter data, wherein the second rock mass parameter data is used to represent the rock mass parameter to be predicted the data of the second process prediction parameter; and

The result module 63 is configured to perform verification processing on the second rock mass parameter data to obtain target rock mass parameter data.

The specific manner of performing operations of each unit in the above embodiment has been described in detail in the embodiment of the method, and will not be described in detail here.

To sum up, in this application, the excavation characteristic data is obtained by preprocessing the acquired data to be processed based on data analysis; the rock mass parameter prediction is performed on the excavation characteristic data based on the preset rock mass parameter prediction model processing to obtain the parameter data of the target rock mass. By processing and filtering the excavation characteristic data, the rock mass parameters are predicted by the rock mass parameter prediction model, and the target rock mass parameter data is obtained, which solves the problem that the accuracy of rock mass parameter judgment in the TBM tunnel construction process in the prior art is relatively low. Low technical problems, and achieve the technical effect of improving the accuracy of rock mass parameter prediction.

It should be noted that the steps shown in the flowcharts of the accompanying drawings may be performed 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 may be performed in an order different than here.

Obviously, those skilled in the art should understand that each unit or each step of the above-mentioned application can be realized by a general-purpose computing device, and they can be concentrated on a single computing device, or distributed in a network composed of multiple computing devices Optionally, they can be implemented with program codes executable by a computing device, so that they can be stored in a storage device and executed by a computing device, or they can be made into individual integrated circuit modules, or they can be integrated into Multiple modules or steps are fabricated into a single integrated circuit module to realize. As such, the present application is not limited to any specific combination of hardware and software.

The above descriptions are only preferred embodiments of the present application, and are not intended to limit the present application. For those skilled in the art, there may be various modifications and changes in the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of this application shall be included within the protection scope of this application.

Claims (10)

1. A data processing method for rock mass parameter prediction is characterized by comprising the following steps:

acquiring data to be processed, wherein the data to be processed is related data used for representing the rock mass tunneling construction of tunneling equipment to be predicted;

preprocessing the data to be processed based on data analysis to obtain tunneling characteristic data, wherein the tunneling characteristic data is used for representing the tunneling characteristics of the tunneling equipment; and

and performing rock mass parameter prediction processing on the tunneling characteristic data based on a preset rock mass parameter prediction model to obtain target rock mass parameter data.

2. The data processing method according to claim 1, wherein rock mass parameter prediction processing is performed on the tunneling characteristic data based on a preset rock mass parameter prediction model, and obtaining target rock mass parameter data comprises:

performing first prediction processing on the tunneling characteristic data based on a preset first rock mass parameter model to obtain first rock mass parameter data, wherein the first rock mass parameter data is used for representing data of a first process prediction parameter of the rock mass to be predicted;

performing characteristic analysis processing on the tunneling characteristic data and the first rock mass parameter data to obtain process characteristic data;

performing second prediction processing on the process characteristic data based on a preset second rock mass parameter model to obtain second rock mass parameter data, wherein the second rock mass parameter data is used for representing data of a second process prediction parameter of the rock mass parameter to be predicted; and

and carrying out verification processing on the second rock mass parameter data to obtain the target rock mass parameter data.

3. The data processing method according to claim 2, wherein the characteristic analysis processing is performed on the tunneling characteristic data and the first rock mass parameter data, and the obtaining of the process characteristic data comprises:

identifying the rock integrity characteristics of the first rock parameters to obtain rock integrity characteristic data, wherein the rock integrity characteristic data is used for representing the rock integrity characteristics to be predicted;

carrying out rock integrity classification prediction processing on the rock integrity characteristic data to obtain classification characteristic data, wherein the classification characteristic data comprises rock integrity classes and class probabilities corresponding to the rock integrity classes; and

and matching the classified characteristic data with the tunneling characteristic data to obtain the process characteristic data.

4. The data processing method according to claim 1, wherein the preprocessing of the data to be processed based on data analysis to obtain the tunneling characteristic data comprises:

acquiring reference data, wherein the reference data is related data of tunneling under the no-load state of the tunneling equipment;

filtering the data to be processed based on the reference data to obtain process data to be processed, wherein the process data to be processed is the data to be processed after the reference data is filtered; and

and performing feature extraction processing on the process data to be processed based on a preset feature extraction rule to obtain the tunneling feature data, wherein the preset feature extraction rule corresponds to the tunneling feature, and the tunneling feature data is the data of the tunneling feature extracted from the process data to be processed.

5. The data processing method of claim 1, wherein prior to obtaining the data to be processed, the method further comprises:

acquiring training sample data, wherein the training sample data is the sample data for training the preset rock mass parameter prediction model;

preprocessing the training sample data based on data analysis to obtain a plurality of training characteristic data, wherein the training characteristic data are a plurality of characteristic data associated with rock mass parameters; and

and training a preset neural network model based on the plurality of training characteristic data to obtain the preset rock mass parameter prediction model.

6. The data processing method of claim 5, wherein training a preset neural network model based on the training feature data to obtain the preset rock mass parameter prediction model comprises:

classifying the plurality of training feature data to obtain a first training feature data set and a second training feature data set, wherein the first training feature data set comprises a plurality of first training feature data, and the second training feature data set comprises a plurality of second training feature data;

training a preset regression classification model according to the first training characteristic data set to obtain a first prediction model of the rock mass, wherein the first prediction model of the rock mass is used for classifying the integrity of the rock mass;

performing rock integrity characteristic prediction processing on the second training characteristic data set based on the first rock mass prediction model to obtain process training characteristic data, wherein the process training characteristic data are characteristic data used for representing rock integrity; and

and training a preset neural network model according to the second training characteristic data set and the process training characteristic data to obtain the rock mass parameter prediction model.

7. A data processing apparatus for rock mass parameter prediction, comprising:

the device comprises a data acquisition module, a data prediction module and a data prediction module, wherein the data acquisition module is used for acquiring data to be processed, and the data to be processed is related data used for representing the tunneling construction of tunneling equipment on a rock mass to be predicted;

the preprocessing module is used for preprocessing the data to be processed based on data analysis to obtain tunneling characteristic data, wherein the tunneling characteristic data is used for representing the tunneling characteristics of the tunneling equipment; and

and the prediction module is used for carrying out rock parameter prediction processing on the tunneling characteristic data based on a preset rock parameter prediction model to obtain target rock parameter data.

8. The data processing apparatus of claim 7, wherein the prediction module comprises:

the first prediction module is used for performing first prediction processing on the tunneling characteristic data based on a preset first rock mass parameter model to obtain first rock mass parameter data, wherein the first rock mass parameter data is used for representing data of a first process prediction parameter of the rock mass to be predicted;

performing characteristic analysis processing on the tunneling characteristic data and the first rock mass parameter data to obtain process characteristic data;

the second prediction module is used for performing second prediction processing on the process characteristic data based on a preset second rock mass parameter model to obtain second rock mass parameter data, wherein the second rock mass parameter data are used for representing data of a second process prediction parameter of the rock mass parameter to be predicted; and

and the result module is used for verifying the second rock mass parameter data to obtain the target rock mass parameter data.

9. A computer readable storage medium storing computer instructions for causing a computer to perform the data processing method for rock mass parameter prediction according to any one of claims 1 to 6.

10. An electronic device, comprising: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executable by the at least one processor to cause the at least one processor to perform a data processing method for rock mass parameter prediction according to any one of claims 1 to 6.

Images