Method and device for predicting water inrush of Ordovician aquifer in coal mining

Abstract

The embodiment of the invention provides a method and a device for predicting water inrush of an Ordovician aquifer in coal mining. The method comprises the following steps: acquiring geological and hydrological data of an Ordovician aquifer, preprocessing the geological and hydrological data, and dividing the geological and hydrological data into a training set and a test set; constructing an XGBoost model, and defining hyper-parameters, a search space of the hyper-parameters and a physical constraint loss function; performing hyper-parameter tuning on the hyper-parameter search space to obtain an optimal hyper-parameter combination; the XGBoost model is trained through a training set, and a trained XGBoost model is obtained; and inputting to-be-predicted data into the trained XGBoost model, and outputting a prediction result. In this way, the prediction process can be clearly presented, the probability of water inrush of each monitoring point can be output and graded, and therefore the interpretability and practicability of the prediction result are improved.

Description

Technical Field

[0001] The present invention generally relates to the technical field of safe coal mining, and more specifically, to a method and device for predicting water inrush events in an Ordovician aquifer during coal mining. Background Art

[0002] The problem of water inrush from the floor of coal mines is a key issue that affects the smooth operation of coal mining and is a core scientific issue in the field of coal mine safety mining technology. With the progress of deep coal mining, the threat of water inrush from the floor of the Ordovician aquifer to coal mining is increasing. However, due to the complexity of the on-site environment, it is extremely difficult to directly obtain the laws of water inrush. Therefore, how to accurately and quickly predict the occurrence time of water inrush in coal mines and provide an effective reference for timely warning and long-term mining of mining projects has important practical and theoretical value.

[0003] In the process of coal mining, traditional water inrush prediction methods rely on simple geological and hydrological monitoring and linear formulas, ignoring the influence of complex geological conditions, resulting in insufficient prediction accuracy. At present, the water inrush coefficient method is mainly used to assess the danger of floor water inrush, but due to the complexity of geological conditions, this method is difficult to achieve accurate prediction. Traditional methods rely too much on manual experience and static historical data, lack real-time performance, and feature selection is full of subjectivity, making the model not widely applicable. In addition, traditional detection methods require frequent drilling and are costly. In contrast, machine learning technology has shown advantages in water inrush prediction due to its ability to process high-dimensional nonlinear data, but existing machine learning prediction of coal mine floor water inrush research can only simply predict whether there is water inrush, and cannot output specific water inrush probability and risk level classification. Summary of the invention

[0004] According to an embodiment of the present invention, a prediction scheme for water inrush in the Ordovician aquifer during coal mining is provided. This scheme can output the probability of water inrush at the monitoring point, thereby prejudging the risk, solving the problem that the current prediction of water inrush in the coal mine floor through machine learning cannot output the specific probability of water inrush, resulting in the inability to clarify the risk level.

[0005] In a first aspect of the present invention, a method for predicting water inrush from an Ordovician aquifer during coal mining is provided. The method comprises:

[0006] Acquire geological and hydrological data of the Ordovician aquifer, preprocess the geological and hydrological data, and divide the preprocessed data into a training set and a test set, wherein the test set includes the data to be predicted;

[0007] Build an XGBoost model, define hyperparameters of the XGBoost model, a search space for hyperparameters, and a physical constraint loss function;

[0008] Use Bayesian optimization method to tune hyperparameters in hyperparameter search space and obtain the optimal hyperparameter combination;

[0009] The XGBoost model is trained by using a training set to obtain a trained XGBoost model;

[0010] The data to be predicted is input into the trained XGBoost model, and a prediction result is output, wherein the prediction result indicates the probability of a water inrush event.

[0011] Furthermore, the geological and hydrological data include drilling data, pressure of aquifers, fault information and water-rich characteristics.

[0012] Furthermore, the hyperparameters include: learning rate, maximum depth of the tree, number of trees, subsample ratio, column sampling ratio, L1 regularization parameter and L2 regularization parameter; wherein

[0013] The search space of learning rate is 0.01~0.1; the search space of maximum tree depth is 3~14; the search space of subsample ratio is 0.5~1; the search space of column sampling ratio is 0.5~1; the search space of L1 regularization parameter is 0~1; the search space of L2 regularization parameter is 0.000001~1.0.

[0014] Furthermore, the physical constraint loss function of the XGBoost model is a combination of the cross entropy loss function and the physical constraint term, including:

[0015]

[0016] Where L is the physical constraint loss function of the XGBoost model; n is the number of samples; i is the sample; is the cross entropy loss function; ‖q i -q pred，i ‖ 2 is a physical constraint term, which is used to measure the loss term of the difference between the flow velocity predicted by the XGBoost model and the actual measured flow velocity; i is the actual measured flow rate; q pred，i is the flow rate predicted by the XGBoost model; λ is a tuning parameter used to adjust the importance of the physical constraint term.

[0017] Furthermore, the cross entropy loss function is:

[0018]

[0019] in, is the cross entropy loss function; C is the number of categories; j is the category; y ij is the label of the i-th sample belonging to the j-th category; The probability that the i-th sample belongs to the j-th class predicted by the XGBoost model.

[0020] Furthermore, it also includes:

[0021] A risk threshold is set, where the risk threshold is between 0 and 1 and is used to distinguish the risk level of a water inrush event.

[0022] Furthermore, the risk level of water inrush events is differentiated by risk thresholds, including:

[0023] If the first prediction result is greater than the risk threshold and the second prediction result is not greater than the risk threshold, the risk level corresponding to the first prediction result is higher than the risk level corresponding to the second prediction result.

[0024] In a second aspect of the present invention, a device for predicting water inrush from an Ordovician aquifer during coal mining is provided. The device comprises:

[0025] A data processing module, used for obtaining geological and hydrological data of the Ordovician aquifer, preprocessing the geological and hydrological data, and dividing the preprocessed data into a training set and a test set, wherein the test set includes data to be predicted;

[0026] A model building module, used to build an XGBoost model, define the hyperparameters of the XGBoost model, the search space of the hyperparameters, and the physical constraint loss function;

[0027] The tuning module is used to use the Bayesian optimization method to perform hyperparameter tuning on the hyperparameter search space to obtain the optimal hyperparameter combination;

[0028] A training module, used to train the XGBoost model using a training set to obtain a trained XGBoost model;

[0029] The prediction module is used to input the data to be predicted into the trained XGBoost model and output the prediction results.

[0030] In a third aspect of the present invention, an electronic device is provided. The electronic device has at least one processor; and a memory connected to the at least one processor in communication; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the method of the first aspect of the present invention.

[0031] In a fourth aspect of the present invention, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute the method of the first aspect of the present invention.

[0032] It should be understood that the contents described in the summary of the invention are not intended to limit the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The above and other features, advantages and aspects of the embodiments of the present invention will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, wherein:

[0034] Figure 1 A flow chart showing a method for predicting water inrush from an Ordovician aquifer during coal mining according to an embodiment of the present invention is shown;

[0035] Figure 2 A block diagram of a device for predicting water inrush from an Ordovician aquifer during coal mining according to an embodiment of the present invention is shown;

[0036] Figure 3 A block diagram of an exemplary electronic device capable of implementing embodiments of the present invention is shown;

[0037] Among them, 300 is an electronic device, 301 is a computing unit, 302 is a ROM, 303 is a RAM, 304 is a bus, 305 is an I / O interface, 306 is an input unit, 307 is an output unit, 308 is a storage unit, and 309 is a communication unit. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] In addition, the term "and / or" in this article is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0040] Figure 1 A flow chart of a method for predicting water inrush from an Ordovician aquifer during coal mining according to an embodiment of the present invention is shown.

[0041] The method includes:

[0042] S101, obtaining geological and hydrological data of the Ordovician aquifer, preprocessing the geological and hydrological data, and dividing the preprocessed data into a training set and a test set, wherein the test set includes data to be predicted.

[0043] In this embodiment, the geological and hydrological data include, but are not limited to, drilling data, pressure of aquifers, fault information, water richness and other key features.

[0044] In this embodiment, considering the data imbalance problem, the geological and hydrological data need to be preprocessed. Specifically, the adaptive SMOTE algorithm is used to preprocess the geological and hydrological data to solve the data imbalance problem, ensure the quality and quantity of the model training data, and improve the stability and prediction performance of the model.

[0045] As an embodiment of the present invention, the train_test_split function is used to divide the data set into a training set and a test set in a ratio of 8:2, and random_state=42 is set to ensure consistency of data division during each run, which helps to maintain the repeatability of the results.

[0046] As an embodiment of the present invention, in order to reveal the key features and their interactions that affect water inrush, SHAP (SHapley Additive exPlanations) values ​​are used.

[0047] Specifically, first, the SHAP library and other necessary libraries were installed and imported, the trained model was loaded, and the feature names were defined. Next, a SHAP interpreter object was created, and the test data was interpreted to calculate the SHAP value of each test sample. By generating summary graphs and force graphs of SHAP values, it is intuitively shown which features have the greatest impact on the model prediction, and the specific contribution direction (positive or negative) of each feature to the prediction results of a specific sample is analyzed.

[0048] Through the analysis of SHAP values, the key features that have the greatest impact on water inrush risk were identified, and a more effective risk management strategy was developed, which provided a new perspective for water inrush risk prevention and control, deepened the understanding of risk factors, ensured the reliability and effectiveness of the model in the Ordovician aquifer water inrush prediction system in coal mining, and provided a reference feature weight in the subsequent optimization and update of the model, which facilitated manual regulation. In addition, SHAP values ​​also revealed the interactions between different features, helping to understand the impact of feature combinations on prediction results.

[0049] In some embodiments, a data management system is also established to store and manage growing data sets. The system includes the use of a distributed file system (Hadoop HDFS) to ensure the efficient storage of large-scale data sets. To ensure data security, data is automatically backed up to a local server every day, and the backup data is uploaded to a remote data center every week. In addition, strict data access rights control is implemented, using identity authentication (LDAP) and access control lists (ACLs) to ensure that only authorized personnel can access sensitive data.

[0050] To improve the predictive power of the model, data preprocessing and feature engineering are performed regularly.

[0051] Specifically, the Python Pandas library is used to check and remove invalid records or malformed data to ensure data quality. Duplicate records are identified and deleted using SQL database query statements or the drop_duplicates method of the Python Pandas library to avoid introducing bias during training.

[0052] For missing values, interpolation and mean are used to fill in the data to maintain data integrity. For discrete variables, mode is used for filling. According to business needs and domain knowledge, geological structure characteristics (fault displacement, coal seam conditions) are extracted from drilling data, and head pressure change characteristics (water pressure fluctuation amplitude, water content change) are extracted from hydrological data.

[0053] S102: Build an XGBoost model, define hyperparameters of the XGBoost model, a search space for the hyperparameters, and a physical constraint loss function.

[0054] The XGBoost model is an optimized gradient boosting decision tree algorithm model. Based on GBDT, it makes improvements in the second-order Taylor expansion of the loss function, the addition of regularization terms, parallel computing, and missing value processing.

[0055] The hyperparameters of the XGBoost model include: learning rate, maximum tree depth, number of trees, subsample ratio, column sampling ratio, L1 regularization parameter and L2 regularization parameter; the search space of hyperparameters is: the search space of learning rate is 0.01~0.1; the search space of maximum tree depth is 3~14; the search space of subsample ratio is 0.5~1; the search space of column sampling ratio is 0.5~1; the search space of L1 regularization parameter is 0~1; the search space of L2 regularization parameter is 0.000001~1.0.

[0056] As an embodiment of the present invention, when initializing the XGBoost classifier, the multi-classification task parameter objective = 'multi:softmax' and the number of categories are set to 2, and the number of categories and the evaluation index are specified as the multi-classification logarithmic loss (mlogloss). The custom objective function custom_objective = custom_loss is set by the set_params method.

[0057] In this embodiment, Darcy's Law is used to describe groundwater flow, and the formula is:

[0058]

[0059] where q is the velocity vector, K is the permeability, is the gradient of water head height.

[0060] Near faults and collapse columns, the water head height may jump, resulting in significant differences in water pressure on both sides of the fault; while in the undisturbed state, the initial water head height near faults and collapse columns reflects the distribution of groundwater in the natural state. In order to measure the gap between the model prediction and the actual label, a cross entropy loss function is defined, the formula is:

[0061]

[0062] in, is the cross entropy loss function; C is the number of categories; j is the category; y ij is the label that the i-th sample belongs to the j-th class (usually one-hot encoding); The probability that the i-th sample belongs to the j-th class predicted by the XGBoost model.

[0063] In addition, a physical constraint is defined to ensure that the difference between the predicted flow rate and the actual measured flow rate is minimized, as follows:

[0064]

[0065] Among them, Physical loss is a loss term used to measure the difference between the physical quantity (flow rate) predicted by the model and the actual measured physical quantity; q i is the actual measured flow rate; q pred，i is the flow rate predicted by the XGBoost model; n is the number of samples.

[0066] Combining the cross entropy loss function with the physical constraint term, we get the loss function:

[0067]

[0068] Where L is the physical constraint loss function of the XGBoost model; n is the number of samples; i is the sample; is the cross entropy loss function; ‖q i -q pred，i ‖ 2 is a physical constraint term, which is used to measure the loss term of the difference between the flow velocity predicted by the XGBoost model and the actual measured flow velocity; i is the actual measured flow rate; q pred，i is the flow rate predicted by the XGBoost model; λ is a tuning parameter used to adjust the importance of the physical constraint term.

[0069] The beneficial effects of setting the physical constraint loss function By ensuring that the physical phenomenon that the increase in the seepage rate of water near the fault and the collapse column leads to an increase in the probability of water inrush in the coal mine, the prediction results of the model will be more consistent with the actual physical laws. In the prediction of coal mine water inrush, the increase in permeability at the fault and collapse column means that the probability of water inrush in the coal mine will also increase. Adding this constraint to the physical constraint loss function can ensure that the model's prediction does not violate this physical phenomenon, thereby improving the credibility of the model. In addition, the physical constraint, as an additional supervisory signal, helps the model better capture the intrinsic structure in the data, reduce the risk of overfitting, and improve the generalization ability of the model on unknown data. In this way, the model not only learns based on data, but also follows known physical laws, which enhances the interpretability and robustness of the model, making it more reliable and practical in practical applications.

[0070] S103. Use the Bayesian optimization method to perform hyperparameter tuning on the hyperparameter search space to obtain the optimal hyperparameter combination.

[0071] In this embodiment, a Bayesian optimization method is used to perform hyperparameter tuning, and the range of hyperparameters that need to be optimized and their prior distribution are clarified.

[0072] Specifically, first, some initial hyperparameter combinations are selected and their performance in the real model is evaluated. These initial data points are used to build a Gaussian Process Model that can predict the possible performance of other hyperparameter combinations and estimate the uncertainty of the prediction. The Gaussian Process Model is used to predict the performance of different hyperparameter combinations and calculate the predicted value and uncertainty of each combination. The Probability of Improvement (PI) strategy is used to select the next hyperparameter combination to be evaluated. The PI strategy selects combinations with good prediction performance and high uncertainty, which means that these combinations have great potential to become better choices. After selecting a new hyperparameter combination, its performance is evaluated in the real model to generate new evaluation results. The new evaluation results are added to the existing dataset and used to update the Gaussian Process Model. By evaluating new hyperparameter combinations, more information can be obtained, thereby improving the prediction ability of the surrogate model. The Gaussian Process Model is updated with new data points to make the model more accurately reflect the actual situation. The updated model will better predict the performance of other hyperparameter combinations. As the iteration proceeds, the prediction accuracy of the model continues to improve, making it possible to more accurately select the next hyperparameter combination to be evaluated. The above process will be repeated many times until the predetermined stopping criteria (such as the maximum number of iterations) are reached. Each iteration will select new candidate points for evaluation and update the proxy model to gradually approach the optimal hyperparameter combination. Through continuous iteration and optimization, the performance of the model can be gradually improved and the best hyperparameter configuration can be found. In the Bayesian optimization process, 50 iterations are set, 5-fold cross validation is used, accuracy and AUC values ​​are used as scoring criteria, and all available processors are used to accelerate calculations.

[0073] As an implementation of the above embodiment, through multiple rounds of experimental verification, the optimal hyperparameter configuration is determined as follows: the learning rate learning_rate is 0.2227, the maximum depth max_depth is 5, the number of trees n_estimators is 294, and the L2 regularization strength parameter reg_lambda is set to 1. This configuration significantly improves the prediction performance and accuracy of the model.

[0074] S104, training the XGBoost model using the training set to obtain a trained XGBoost model. The training process is implemented based on the physical constraint loss function of the XGBoost model.

[0075] S105, inputting the data to be predicted into the trained XGBoost model, and outputting a prediction result, wherein the prediction result indicates the probability of a water inrush event.

[0076] In this example, the SMOTE-PIBO-XGB (5-fold Cross-Validation) model constructs 294 decision trees during the training phase. Each tree is optimized based on the residual between the prediction result of the previous tree and the actual result, with the goal of minimizing the loss function. In this process, the model is actually adjusting the logarithmic probability, which is defined as the natural logarithm of the ratio of the probability of an event occurring to the probability of not occurring:

[0077]

[0078] Among them, log odds is the natural logarithm of the ratio of the probability of an event occurring to the probability of not occurring, and p represents the probability of a water inrush event occurring at a certain location.

[0079] Second, to convert the log odds into easy-to-understand probability values, the XGBoost model uses the Sigmoid function (also known as the Logistic function) to perform the conversion:

[0080]

[0081] Where p' represents the probability value after transformation. x represents the logarithmic probability. When x = 0, p' = 0.5.

[0082] Through the Sigmoid function, a value between 0 and 1 can be obtained, which can intuitively reflect the possibility of water inrush at the location. If the calculated probability is close to 1, it means that the risk of water inrush at the location is very high; conversely, if the probability is close to 0, it means that the risk of water inrush at the location is relatively low.

[0083] By quantifying the risk of water inrush into a probability value, the probability of water inrush is represented by the size of the output probability value. The larger the output probability value, the greater the probability of water inrush, and the smaller the output probability of water inrush, the lower the probability of water inrush. The output of water inrush probability can assist decision makers in conducting more scientific risk management.

[0084] In this embodiment, a risk threshold value may also be set, and the risk threshold value is between 0 and 1, and the risk level of water inrush events may be distinguished by the risk threshold value. Specifically, one or more risk threshold values ​​between 0 and 1 may be set. If the first prediction result is greater than the risk threshold value, and the second prediction result is not greater than the risk threshold value, the risk level corresponding to the first prediction result is higher than the risk level corresponding to the second prediction result. Risk level situations may also be visually partitioned.

[0085] As an implementation method of the above embodiment, the thresholds of the four risk stages are first set according to the predicted probability, and the risk level layers represented by different colors are created using the GIS platform. Then, a user-friendly front-end interface is designed so that users can easily view and obtain information about the required area. When the user clicks on any area on the map, the system automatically sends a request to the back-end server, which extracts relevant data from the database and returns it to the front-end. Finally, the front-end will pop up an information box on the map, displaying the risk level, predicted probability and other important information of the clicked area, realize the automatic division and visual zoning of water inrush risk, help users quickly grasp key details, and improve risk management efficiency and emergency response capabilities. With the help of the geographic information system (GIS), these probability data are further visualized, and the water inrush risk of each area is intuitively evaluated, providing a more scientific and accurate basis for the safety management and risk prevention of coal mines. In this way, managers can more effectively identify high-risk areas and take targeted measures to ensure the safety of mine operations.

[0086] In some embodiments, the model performance is evaluated by 5-fold cross validation to ensure that the model not only has a good performance on the training set, but also has a strong generalization ability on the test set.

[0087] Through the accumulated actual operation data, the parameter configuration of the model is continuously adjusted and optimized. Specifically, after the system is built, the system will automatically collect feedback data from actual application scenarios and use it in the model retraining process. By learning from a large amount of data, the model can self-adjust weights and other parameters, thereby gradually improving its prediction accuracy for unknown data.

[0088] In order to avoid the need to train the model from scratch every time new data is obtained, incremental learning technology is used. The specific implementation method is to save the intermediate model state (such as tree structure, weight, etc.). When new data is received, only incremental updates need to be made based on the existing model, without the need to train from scratch, saving a lot of time and computing resources.

[0089] In addition, online learning is implemented, allowing the model to learn new data in real time or near real time to adapt to changes in data patterns. The specific implementation method is to process new data in real time through a streaming processing framework (Apache Flink) and update the model state.

[0090] In some optional implementations of this embodiment, the model is evaluated and verified regularly, and the model performance is regularly evaluated through indicators such as accuracy and F1 score, and its generalization ability on new data is ensured. The effects of different versions of the model are compared by testing and other means, and the best version is selected to be deployed in the production environment. The trained model is deployed to the production environment and integrated with the existing business system to ensure that the model can process new data in real time. The operation status of the model is continuously monitored, and logging and alarm mechanisms (such as Prometheus and Alertmanager) are used to ensure the stability of the model's performance and to detect any abnormalities in a timely manner. Not only does it improve the predictive performance of the model, but it also enhances its universality and practicality in practical applications, ensuring the reliability and effectiveness of the model in the Ordovician aquifer water inrush prediction system in coal mining.

[0091] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described order of actions, because according to the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0092] The above is an introduction to a method embodiment. The following is a further explanation of the scheme of the present invention through a device embodiment having the same inventive concept as the method in the aforementioned embodiment.

[0093] like Figure 2 As shown, the device 200 includes:

[0094] The data processing module 210 is used to obtain geological and hydrological data of the Ordovician aquifer, pre-process the geological and hydrological data, and divide the pre-processed data into a training set and a test set, wherein the test set includes the data to be predicted.

[0095] The model building module 220 is used to build an XGBoost model, define the hyperparameters of the XGBoost model, the search space of the hyperparameters, and the physical constraint loss function.

[0096] The tuning module 230 is used to perform hyperparameter tuning on the hyperparameter search space using the Bayesian optimization method to obtain the optimal hyperparameter combination.

[0097] The training module 240 is used to train the XGBoost model using a training set to obtain a trained XGBoost model.

[0098] The prediction module 250 is used to input the data to be predicted into the trained XGBoost model and output the prediction result.

[0099] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the described module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0100] In the technical solution of the present invention, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0101] According to an embodiment of the present invention, the present invention further provides an electronic device and a readable storage medium.

[0102] Figure 3 A schematic block diagram of an electronic device 300 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0103] The electronic device 300 includes a computing unit 301, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 302 or a computer program loaded from a storage unit 308 into a random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 can also be stored. The computing unit 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0104] A number of components in the electronic device 300 are connected to the I / O interface 305, including: an input unit 306, such as a keyboard, a mouse, etc.; an output unit 307, such as various types of displays, speakers, etc.; a storage unit 308, such as a disk, an optical disk, etc.; and a communication unit 309, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 309 allows the electronic device 300 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0105] The computing unit 301 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 301 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 301 performs the various methods and processes described above, such as methods S101 to S105. For example, in some embodiments, methods S101 to S105 may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 308. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 300 via the ROM 302 and / or the communication unit 309. When the computer program is loaded into the RAM 303 and executed by the computing unit 301, one or more steps of the methods S101 to S105 described above may be performed. Alternatively, in other embodiments, the computing unit 301 may be configured to execute methods S101 to S105 in any other appropriate manner (eg, by means of firmware).

[0106] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0107] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0108] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0109] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0110] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0111] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0112] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0113] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for predicting water inrush from an Ordovician aquifer during coal mining, characterized in that: include: Acquire geological and hydrological data of the Ordovician aquifer, preprocess the geological and hydrological data, and divide the preprocessed data into a training set and a test set, wherein the test set includes the data to be predicted; Build an XGBoost model, define hyperparameters of the XGBoost model, a search space for the hyperparameters, and a physical constraint loss function; Use Bayesian optimization method to tune hyperparameters in hyperparameter search space and obtain the optimal hyperparameter combination; The XGBoost model is trained by using a training set to obtain a trained XGBoost model; The data to be predicted is input into the trained XGBoost model, and a prediction result is output, wherein the prediction result indicates the probability of a water inrush event.

2. The method according to claim 1, characterized in that The geological and hydrological data include borehole data, aquifer pressure, fault information and water-bearing characteristics.

3. The method according to claim 1, characterized in that The hyperparameters include: learning rate, maximum depth of tree, number of trees, subsample ratio, column sampling ratio, L1 regularization parameter and L2 regularization parameter; The search space of learning rate is 0.01~0.1; the search space of maximum tree depth is 3~14; the search space of subsample ratio is 0.5~1; the search space of column sampling ratio is 0.5~1; the search space of L1 regularization parameter is 0~1; the search space of L2 regularization parameter is 0.000001~1.

0.

4. The method according to claim 1, characterized in that: The physical constraint loss function of the XGBoost model is a combination of the cross entropy loss function and the physical constraint term, including: Where L is the physical constraint loss function of the XGBoost model; n is the number of samples; i is the sample; is the cross entropy loss function; ‖q i -q pred，i ‖ 2 is a physical constraint term, which is used to measure the loss term of the difference between the flow velocity predicted by the XGBoost model and the actual measured flow velocity; i is the actual measured flow rate; q pred，i is the flow rate predicted by the XGBoost model; λ is a tuning parameter used to adjust the importance of the physical constraint term.

5. The method according to claim 4, characterized in that The cross entropy loss function is: in, is the cross entropy loss function; C is the number of categories; j is the category; y ij is the label of the i-th sample belonging to the j-th category; The probability that the i-th sample belongs to the j-th class predicted by the XGBoost model.

6. The method according to claim 1, characterized in that Also includes: A risk threshold is set, where the risk threshold is between 0 and 1 and is used to distinguish the risk level of a water inrush event.

7. The method according to claim 6, characterized in that The risk level of water inrush events is differentiated by risk thresholds, including: If the first prediction result is greater than the risk threshold and the second prediction result is not greater than the risk threshold, the risk level corresponding to the first prediction result is higher than the risk level corresponding to the second prediction result.

8. A device for predicting water inrush from an Ordovician aquifer during coal mining, characterized in that: include: A data processing module, used for obtaining geological and hydrological data of the Ordovician aquifer, preprocessing the geological and hydrological data, and dividing the preprocessed data into a training set and a test set, wherein the test set includes data to be predicted; A model building module, used to build an XGBoost model, define the hyperparameters of the XGBoost model, the search space of the hyperparameters, and the physical constraint loss function; The tuning module is used to use the Bayesian optimization method to perform hyperparameter tuning on the hyperparameter search space to obtain the optimal hyperparameter combination; A training module, used to train the XGBoost model using a training set to obtain a trained XGBoost model; The prediction module is used to input the data to be predicted into the trained XGBoost model and output the prediction results.

9. An electronic device comprising at least one processor; and a memory connected in communication with the at least one processor; characterized in that: The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-7.

Images