Drainage system
By constructing a recurrent neural network model based on historical precipitation data, the problem of inaccurate prediction of drainage volume in coal mine water storage tanks was solved, and precise drainage control was achieved.
Patent Information
- Application Number
- CN202511490418.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-17
- Publication Date
- 2026-01-13
AI Technical Summary
The existing drainage systems of coal mine water sump systems cannot accurately obtain drainage volume related to precipitation, resulting in inaccurate drainage.
By constructing precipitation sequence data based on historical precipitation data from mine water reservoirs, a recurrent neural network model is used to train a drainage prediction module, which is then combined with multi-stage centrifugal pumps for precise drainage.
This improves the accuracy of drainage volume prediction, ensuring the accuracy and efficiency of mine water tank drainage.
Smart Images
Figure CN121322091A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of tunnel drainage technology, and in particular to a drainage system. Background Technology
[0002] Coal mine drainage systems are a crucial part of coal mine production, directly affecting the safety of mine roadways. With the expansion of coal mining scale and the increase in mine depth, the design of mine drainage systems has become increasingly complex and important. Traditional mine drainage systems mainly rely on centrifugal pumps and submersible pumps, which show significant inadequacy in adapting to complex mine hydrological environments. Currently, most coal mines still use traditional water level monitoring drainage methods. This method determines the drainage volume of the mine water tank based on the water level in the mine water tank using manual or semi-manual methods. However, the drainage volume in the mine water tank is usually related to precipitation, and this method cannot accurately obtain the drainage volume related to precipitation, leading to technical problems of inaccurate drainage when draining the mine water tank based on the drainage volume. Summary of the Invention
[0003] In view of this, the purpose of the present invention is to provide a drainage system that can solve the technical problem in the prior art that the drainage volume related to precipitation cannot be accurately obtained, resulting in inaccurate drainage of mine water tanks.
[0004] To achieve the above objectives, the technical solutions adopted in the embodiments of the present invention are as follows: In a first aspect, embodiments of the present invention provide a drainage system, wherein the system includes: Data acquisition module: used to construct precipitation sequence data based on historical precipitation data from mine water reservoirs; Discharge prediction module: used to obtain the hidden state of precipitation corresponding to the precipitation sequence data, and use the hidden state of precipitation as the input of a linear activation function to determine the predicted discharge volume; wherein, the discharge prediction module is trained from a training dataset of precipitation sequence data labeled with discharge volume; Drainage module: used to drain the mine water tank based on the predicted drainage volume.
[0005] Furthermore, this embodiment of the invention provides a first possible implementation of the first aspect, wherein the precipitation sequence data constructed based on historical precipitation data from mine water reservoirs includes: Historical precipitation data of the mine water reservoir is obtained, and the historical precipitation data is preprocessed; wherein, the preprocessing includes interpolation processing and outlier processing; The preprocessed historical precipitation data is standardized to obtain the precipitation sequence data.
[0006] Furthermore, this embodiment of the invention provides a second possible implementation of the first aspect, wherein obtaining the hidden precipitation state corresponding to the precipitation sequence data and using the hidden precipitation state as input to a linear activation function to determine the predicted drainage volume includes: Precipitation sequence data, a preset weight matrix, and a preset bias vector are used as inputs to a nonlinear activation function to determine the precipitation cell state corresponding to the precipitation sequence data. The precipitation cell state is used as the input of the hyperbolic tangent function to determine the precipitation hidden state; The hidden precipitation state is used as input to a linear activation function to determine the predicted drainage volume.
[0007] Furthermore, this embodiment of the invention provides a third possible implementation of the first aspect, wherein the drainage system further includes: a training module; The training module is used to acquire multiple historical precipitation data and corresponding drainage volumes, generate corresponding precipitation sequence data based on each of the historical precipitation data, label each of the precipitation sequence data and the corresponding drainage volumes, and establish a sample dataset based on the labeled precipitation sequence data; wherein, the sample dataset includes the training dataset. The training dataset is input into a recurrent neural network model for training to obtain the water displacement prediction module.
[0008] Furthermore, this embodiment of the invention provides a fourth possible implementation of the first aspect, wherein the drainage system further includes: a performance testing module; The sample dataset also includes a test dataset; The performance testing module is used to input the test dataset into the query retrieval module to obtain performance testing parameters to evaluate the performance of the query retrieval module; wherein, the performance testing parameters include: root mean square error and mean absolute error.
[0009] Furthermore, this embodiment of the invention provides a fifth possible implementation of the first aspect, wherein the drainage system further includes: a human-computer interaction module; The human-computer interaction module is used to display the water level of the mine water in the mine water tank in real time and to allow the user to modify the predicted drainage volume obtained by the drainage volume prediction module.
[0010] Furthermore, this embodiment of the invention provides a sixth possible implementation of the first aspect, wherein the drainage system further includes: a siltation monitoring module; The siltation monitoring module is used to monitor the siltation thickness at the bottom of the mine water tank in real time. When the siltation thickness is greater than a first preset thickness, the module controls the human-machine interface to issue a dredging instruction.
[0011] Furthermore, this embodiment of the invention provides a seventh possible implementation of the first aspect, wherein the drainage system further includes: a water tank dredging module; The water tank dredging module is used to discharge the mud and sand at the bottom of the mine water tank when the human-computer interaction module issues the dredging instruction.
[0012] Furthermore, this embodiment of the invention provides an eighth possible implementation of the first aspect, wherein the drainage system further includes: a water level monitoring module; The water level monitoring module is used to monitor the water level of the mine water in the mine water tank in real time. If the water level is lower than the first preset water level, the predicted drainage volume is reduced; if the water level is higher than the second preset water level, an audible and visual alarm is issued and the predicted drainage volume is increased.
[0013] Furthermore, this embodiment of the invention provides a ninth possible implementation of the first aspect, wherein the drainage system further includes: a water quality detection module; The water quality detection module is used to monitor the pH value of the mine water in the mine water tank in real time. If the pH value is less than a first preset value, an alkaline substance is added to the mine water to increase the pH value. The alkaline substance includes sodium carbonate and sodium hydroxide.
[0014] This invention provides a drainage system comprising: a data acquisition module for constructing precipitation sequence data based on historical precipitation data from a mine water reservoir; a drainage volume prediction module for acquiring the hidden precipitation state corresponding to the precipitation sequence data and using the hidden precipitation state as input to a linear activation function to determine the predicted drainage volume; wherein the drainage volume prediction module is trained using a training dataset of historical precipitation data labeled with drainage volume; and a drainage module for draining water from the mine water reservoir based on the predicted drainage volume. In this invention, the data acquisition module acquires historical precipitation data from the mine water reservoir and constructs precipitation sequence data based on this historical precipitation data. The drainage volume prediction module has been trained using a training dataset of precipitation sequence data labeled with drainage volume. Therefore, inputting the precipitation sequence data obtained from historical precipitation data into the drainage volume prediction module yields the predicted drainage volume, improving the accuracy of the predicted drainage volume. Finally, the drainage module drains water from the mine water reservoir based on the predicted drainage volume, thus improving the accuracy of drainage from the mine water reservoir.
[0015] Other features and advantages of the embodiments of the present invention will be set forth in the following description, or some features and advantages may be inferred from the description or determined without doubt, or may be learned by practicing the techniques described above in the embodiments of the present invention.
[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description
[0017] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0018] Figure 1 A schematic diagram of the main modules of a drainage system provided in an embodiment of the present invention is shown; Figure 2 This diagram illustrates the processing steps of a data acquisition module in a drainage system according to an embodiment of the present invention. Figure 3 The diagram illustrates the processing steps of a drainage volume prediction module in a drainage system provided by an embodiment of the present invention. Figure 4 A schematic diagram of the complete modules of a drainage system provided by an embodiment of the present invention is shown. Detailed Implementation
[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be described below in conjunction with the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.
[0020] This embodiment provides a drainage system; for details, please refer to... Figure 1 The diagram shown illustrates the main modules of a drainage system, which mainly includes: Data acquisition module 11: used to construct precipitation sequence data based on historical precipitation data from mine water reservoirs; Historical precipitation documents near the location of the mine water reservoir are obtained through a distributed crawler architecture. Based on these documents, the required historical precipitation data is extracted. Specifically, when precipitation occurs in the environment of the mine water reservoir, it takes time for rainwater to seep from the ground into the underground reservoir. Therefore, there is a lag between precipitation and the water inflow of the reservoir. Thus, precipitation data from 3 to 7 days before the date on which the drainage volume is to be predicted is typically obtained as historical precipitation data. The collected historical precipitation data is then sorted according to date to form a precipitation sequence data.
[0021] Drainage volume prediction module 12: used to obtain the hidden state of precipitation corresponding to precipitation sequence data, and use the hidden state of precipitation as the input of a linear activation function to determine the predicted drainage volume; wherein, the drainage volume prediction module 12 is trained from a training dataset of precipitation sequence data labeled with drainage volume; The drainage prediction module 12 is obtained by training a recurrent neural network with a training dataset of precipitation sequence data labeled with drainage volume. After the precipitation sequence data is input into the trained recurrent neural network, the forget gate, input gate and output gate in the recurrent neural network obtain the hidden precipitation state corresponding to the precipitation sequence data based on the precipitation sequence data. The fully connected layer in the recurrent neural network processes the hidden precipitation state based on the linear activation function to determine the predicted drainage volume of the mine water tank for the day.
[0022] Drainage module 13: Used to drain the mine water tank based on the predicted drainage volume; The drainage module 13 is equipped with multi-stage centrifugal water pumps. When the multi-stage centrifugal water pumps start, the high-speed rotating impellers apply centrifugal force to the mine water in the mine water tank, so that the water gains kinetic energy and pressure. This allows the predicted drainage volume of mine water to be extracted from the mine water tank and discharged to the outside of the mine roadway through the drainage pipe.
[0023] The data acquisition module in the drainage system provided by this invention acquires historical precipitation data of the mine water tank and constructs precipitation sequence data based on the historical precipitation data. The drainage volume prediction module has been trained using a training dataset of precipitation sequence data labeled with drainage volume. Therefore, by inputting the precipitation sequence data obtained based on historical precipitation data into the drainage volume prediction module, the predicted drainage volume can be obtained, which improves the accuracy of the predicted drainage volume. Finally, the drainage module drains the mine water tank based on the predicted drainage volume, which improves the accuracy of the drainage of the mine water tank.
[0024] In one embodiment, see Figure 2 The diagram illustrates the processing steps of a data acquisition module in a drainage system. This embodiment provides a specific implementation method for constructing precipitation sequence data based on historical precipitation data from a mine water reservoir: S201, Obtain historical precipitation data from the mine water reservoir and preprocess the historical precipitation data; the preprocessing includes interpolation and outlier handling. Acquire historical precipitation data for the environment surrounding the mine water reservoir over the past 3-7 days. Perform interpolation on missing values and outlier handling on outliers in the historical precipitation data. Interpolation includes: linearly fitting precipitation data corresponding to the dates before and after missing values in the historical precipitation data using linear interpolation to obtain the interpolated historical precipitation data. Outlier handling includes: calculating the standard deviation of the historical precipitation data; if the precipitation data exceeds the average value ± three times the standard deviation of the historical precipitation data, it indicates that the current day's precipitation data is an extreme outlier. Remove extreme outliers and generate the interpolated historical precipitation data using linear interpolation.
[0025] S203, the preprocessed historical precipitation data is standardized to obtain precipitation sequence data; The highest and lowest precipitation data for the past 3-7 days are obtained from the preprocessed historical precipitation data. The difference between the highest and lowest precipitation data is calculated to obtain the precipitation data difference value. The difference between the historical precipitation data and the lowest precipitation data for each day is calculated. The difference corresponding to the historical precipitation data for each day is compared with the precipitation data difference value to complete the data standardization processing of historical precipitation data.
[0026] In one embodiment, see Figure 3 The diagram illustrates the processing steps of a drainage volume prediction module in a drainage system. This embodiment provides a specific implementation method for obtaining the hidden precipitation state corresponding to precipitation sequence data and using the hidden precipitation state as input to a linear activation function to determine the predicted drainage volume: S301, using precipitation sequence data, a preset weight matrix, and a preset bias vector as inputs to a nonlinear activation function, determines the precipitation cell state corresponding to the precipitation sequence data; The drainage prediction module 12 processes the precipitation sequence data based on the trained recurrent neural network model to obtain the precipitation cell states corresponding to the precipitation sequence data. Specifically, the recurrent neural network model starts from the date furthest from the date to be measured and processes the precipitation sequence data on a daily basis at each time step. At each time step, the forget gate in the recurrent neural network model determines which information needs to be discarded at this time step based on the precipitation cell states obtained from the previous time step, the first preset weight matrix, and the first preset bias vector. ; in, The information to be discarded is determined by the forget gate at time step t. (*) represents a non-linear activation function (i.e., the Sigmoid function); This is the first preset weight matrix; The cell state represents the precipitation at time step t-1. This represents the precipitation sequence data corresponding to the t-th time step; This is the first preset bias vector; The representative will and To splice; At each time step, the input gate of the recurrent neural network model determines which information needs to be retained in the corresponding precipitation cell state at this time step based on the precipitation cell state obtained in the previous time step, the second preset weight matrix, and the second preset bias vector.
[0027] in, The information to be retained is determined by the input gate at time step t; This is the second preset weight matrix; This is the second preset bias vector; In the recurrent neural network model at each time step, the candidate precipitation cell state for the current time step is determined based on the precipitation cell state obtained from the previous time step, the third preset weight matrix, and the third preset bias vector. ; in, The candidate precipitation cell state generated at time step t; This is the third preset weight matrix; This is the third preset bias vector; Finally, at each time step, the recurrent neural network model determines the precipitation cell state for that time step based on the output of the forget gate, the output of the input gate, and the generated candidate precipitation cell states. ; Among them, the operator This represents multiplying the data step by step (i.e., solving the Hadamard product).
[0028] S303 uses the precipitation cell state as input to the hyperbolic tangent function to determine the precipitation hidden state; At each time step, the output gate of the recurrent neural network model determines the information that needs to be output from the precipitation cell state in the current time step based on the precipitation cell state obtained in the previous time step, the fourth preset weight matrix, and the fourth preset bias vector: ; in, The output information is determined by the output gate at the t-th time step; This is the fourth preset weight matrix; This is the fourth preset bias vector; Ultimately, at each time step, the recurrent neural network model determines the hidden precipitation state based on the output of the output gate and the precipitation cell state at that time step. ; The hidden precipitation state output at the last time step better reflects the overall information of the precipitation sequence data. Therefore, the hidden precipitation state output at the last time step is used as the input to the fully connected layer in the recurrent neural network model.
[0029] S305 uses the hidden precipitation state as input to a linear activation function to determine the predicted drainage volume; In a recurrent neural network model, the linear activation function of the fully connected layer determines the predicted drainage based on the hidden precipitation state output at the last time step.
[0030] In one embodiment, see Figure 4 The diagram shows a complete module of a drainage system. The drainage system provided in this embodiment also includes: a training module 14. The training module 14 is used to acquire multiple historical precipitation data and corresponding drainage volumes, generate corresponding precipitation sequence data based on each historical precipitation data, label each precipitation sequence data and corresponding drainage volumes, and establish a sample dataset based on the labeled precipitation sequence data; wherein, the sample dataset includes the training dataset. Training module 14 acquires long-term historical precipitation data (usually 1-2 years) of the environment where the mine water tank is located. Based on a preset time step, it decomposes the long-term historical precipitation data into multiple historical precipitation data and determines the drainage volume corresponding to each historical precipitation data. Based on each historical precipitation data, it generates its corresponding precipitation sequence data. It then labels each precipitation sequence data with its corresponding drainage volume and establishes a sample dataset based on the labeled precipitation sequence data.
[0031] The training dataset is input into the recurrent neural network model for training, resulting in the displacement prediction module 12. By training a recurrent neural network model using a training dataset, the model can determine the corresponding drainage volume based on precipitation sequence data.
[0032] In one embodiment, such as Figure 4 As shown, the drainage system provided in this embodiment also includes: a performance testing module 15; The sample dataset also includes a test dataset; The performance testing module 15 is used to input the test dataset into the query retrieval module to obtain performance testing parameters to evaluate the performance of the query retrieval module; among which, the performance testing parameters include: root mean square error and mean absolute error; After the test dataset is input into the performance testing module 15, the predicted drainage volume corresponding to each precipitation sequence data is obtained. The root mean square error and mean absolute error are determined based on the predicted drainage volume and the actual drainage volume corresponding to each precipitation sequence data. Specifically, the root mean square error is determined based on the predicted drainage volume and the actual drainage volume corresponding to each precipitation sequence data. ; in, Indicates the root mean square error; This represents the actual drainage volume corresponding to the p-th precipitation sequence data. denoted as the predicted drainage volume corresponding to the p-th precipitation sequence data; n is the total number of precipitation sequence data; the root mean square error is more sensitive to larger errors in the predicted drainage volume, and the smaller the root mean square error value, the better the performance of the trained recurrent neural network model (i.e., drainage volume prediction module 12). The mean absolute error is determined based on the predicted drainage volume and the actual drainage volume corresponding to each precipitation sequence data: ; in, The mean absolute error is represented; the smaller the mean absolute error, the better the performance of the recurrent neural network model (i.e., displacement prediction module 12) after training.
[0033] In one embodiment, such as Figure 4 As shown, the drainage system provided in this embodiment also includes: a human-machine interaction module 16; The human-computer interaction module 16 is used to display the water level of the mine water in the mine water tank in real time and to allow the user to modify the predicted drainage volume obtained by the drainage volume prediction module 12. The human-machine interaction module 16 is equipped with a human-machine interaction interface, specifically including a primary interface and a secondary interface. The primary interface displays a real-time simulation diagram of the entire drainage equipment, including the water tank, all pumps, valves, pipelines, and dynamic display of key parameters (water level, total flow rate, voltage). The operating status of each component in the drainage equipment can be determined based on color: if a component is displayed in green, it indicates that it is in normal operation; if a component is displayed in red, it indicates that it is in a fault state; if a component is displayed in orange, it indicates that it is in a stopped state. The primary interface also allows users to manually modify the predicted drainage volume and adjust the drainage flow rate of the mine water tank in real time. Clicking on the icon of a component in the primary interface will take you to the corresponding secondary interface, which will display the operating parameters of that component.
[0034] In one embodiment, such as Figure 4 As shown, the drainage system provided in this embodiment also includes: a siltation monitoring module 17; The siltation monitoring module 17 is used to monitor the siltation thickness at the bottom of the mine water tank in real time. When the siltation thickness is greater than the first preset thickness, the human-machine interaction module 16 is controlled to issue a dredging instruction. The sediment accumulation monitoring module 17 determines the thickness of the sediment accumulated at the bottom of the mine water tank based on the ultrasonic sediment level gauge installed at the top of the mine water tank. Specifically, since ultrasonic waves propagate at different speeds in water and in sediment, the ultrasonic sediment level gauge releases ultrasonic waves based on this principle to determine the thickness of the sediment accumulation at the bottom of the mine water tank. When the sediment accumulation thickness is greater than the first preset thickness (related to the height of the mine water tank, usually set to 30% of the height of the mine water tank), the human-machine interaction module 16 issues a dredging instruction on the human-machine interaction interface to remind the user that sediment dredging is required.
[0035] In one embodiment, such as Figure 4 As shown, the drainage system provided in this embodiment also includes: a water tank dredging module 18; The water tank dredging module 18 is used to discharge the mud and sand at the bottom of the mine water tank when the human-machine interaction module 16 issues a dredging instruction; Because the silt accumulated at the bottom of the mine water tank can cause wear and tear on the water tank, thereby reducing its service life, the water tank dredging module 18 starts the submersible mud pump to extract and discharge the silt accumulated at the bottom of the mine water tank when the human-machine interaction module 16 issues a dredging instruction. Specifically, when the human-machine interaction module 16 issues a dredging instruction, the water tank dredging module 18 controls the submersible mud pump to reach the silt accumulation position at the bottom of the mine water tank, starts the submersible mud pump, sucks the expected silt from the bottom of the mine water tank into the pump, and discharges it to the outside of the mine water tank through the mud conveying hose.
[0036] In one embodiment, such as Figure 4 As shown, the drainage system provided in this embodiment also includes: a water level monitoring module 19; The water level monitoring module 19 is used to monitor the water level of the mine water tank in real time. If the water level is lower than the first preset water level, the predicted drainage volume is reduced; if the water level is higher than the second preset water level, an audible and visual alarm is issued and the predicted drainage volume is increased. The water level monitoring module 19 monitors the water level in the mine water tank in real time based on water level sensors set at the first and second preset water levels. Specifically, according to the principle of hydrostatics, pressure is directly proportional to liquid level. Therefore, the water level sensor is equipped with a pressure-sensitive diaphragm. If the water level in the mine water tank is lower than the first preset water level, the pressure sensed by the water level sensor at the first preset water level gradually decreases to 0, indicating that the water level in the mine water tank is too low, and the predicted drainage volume can be appropriately reduced. If the water level in the mine water tank is higher than the second preset water level, the pressure sensed by the water level sensor at the second preset water level gradually increases from 0, indicating that the water level in the mine water tank is too high. An audible and visual alarm is issued to remind the staff to avoid the mine water overflowing from the mine water tank, and the predicted drainage volume is immediately and significantly increased, thereby increasing the drainage flow rate in the mine water tank.
[0037] In one embodiment, such as Figure 4 As shown, the drainage system provided in this embodiment also includes: a water quality detection module 20; The water quality detection module 20 is used to monitor the pH value of the mine water in the mine water tank in real time. If the pH value is less than the first preset value, an alkaline substance is added to the mine water to increase the pH value. The alkaline substance includes sodium carbonate and sodium hydroxide. If the mine water is acidic, it will corrode the walls of the mine water tank. Therefore, the water quality detection module 20 samples the mine water in the mine water tank in real time based on the explosion-proof sampling pump and detects the acidity and alkalinity of the mine water through the pH sensor. If the acidity and alkalinity of the mine water is lower than the first preset value (usually set to 6.5), it proves that the mine water is acidic. The neutralizing agent dosing device is activated to add alkaline substances (including sodium carbonate and sodium hydroxide) to the mine water to increase the acidity and alkalinity of the mine water (even if the acidity and alkalinity return to neutral (pH=7)).
[0038] Furthermore, in the description of the embodiments of the present invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention based on the specific circumstances.
[0039] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0040] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.
[0041] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.
Claims
1. A drainage system, characterized in that, include: Data acquisition module: used to construct precipitation sequence data based on historical precipitation data from mine water reservoirs; Discharge prediction module: used to obtain the hidden state of precipitation corresponding to precipitation sequence data, and use the hidden state of precipitation as the input of a linear activation function to determine the predicted discharge volume; wherein, the discharge prediction module is trained from a training dataset of precipitation sequence data labeled with discharge volume; Drainage module: used to drain the mine water tank based on the predicted drainage volume.
2. The drainage system according to claim 1, characterized in that, The precipitation sequence data constructed based on historical precipitation data from mine water reservoirs includes: Historical precipitation data of the mine water reservoir is obtained, and the historical precipitation data is preprocessed; wherein, the preprocessing includes interpolation processing and outlier processing; The preprocessed historical precipitation data is standardized to obtain the precipitation sequence data.
3. The drainage system according to claim 1, characterized in that, The step of obtaining the hidden precipitation state corresponding to the precipitation sequence data and using the hidden precipitation state as the input of a linear activation function to determine the predicted drainage volume includes: Precipitation sequence data, a preset weight matrix, and a preset bias vector are used as inputs to a nonlinear activation function to determine the precipitation cell state corresponding to the precipitation sequence data. The precipitation cell state is used as the input of the hyperbolic tangent function to determine the precipitation hidden state; The hidden precipitation state is used as input to a linear activation function to determine the predicted drainage volume.
4. The drainage system according to claim 1, characterized in that, Also includes: Training module; The training module is used to acquire multiple historical precipitation data and corresponding drainage volumes, generate corresponding precipitation sequence data based on each of the historical precipitation data, label each of the precipitation sequence data and the corresponding drainage volumes, and establish a sample dataset based on the labeled precipitation sequence data; wherein, the sample dataset includes the training dataset. The training dataset is input into a recurrent neural network model for training to obtain the water displacement prediction module.
5. The drainage system according to claim 4, characterized in that, Also includes: Performance testing module; The sample dataset also includes a test dataset; The performance testing module is used to input the test dataset into the query retrieval module to obtain performance testing parameters to evaluate the performance of the query retrieval module; wherein, the performance testing parameters include: root mean square error and mean absolute error.
6. The drainage system according to claim 1, characterized in that, Also includes: Human-computer interaction module; The human-computer interaction module is used to display the water level of the mine water in the mine water tank in real time and to allow the user to modify the predicted drainage volume obtained by the drainage volume prediction module.
7. The drainage system according to claim 6, characterized in that, Also includes: Sedimentation monitoring module; The siltation monitoring module is used to monitor the siltation thickness at the bottom of the mine water tank in real time. When the siltation thickness is greater than a first preset thickness, the module controls the human-machine interface to issue a dredging instruction.
8. The drainage system according to claim 7, characterized in that, Also includes: Water tank dredging module; The water tank dredging module is used to discharge the mud and sand at the bottom of the mine water tank when the human-computer interaction module issues the dredging instruction.
9. The drainage system according to claim 1, characterized in that, Also includes: Water level monitoring module; The water level monitoring module is used to monitor the water level of the mine water in the mine water tank in real time. If the water level is lower than the first preset water level, the predicted drainage volume is reduced; if the water level is higher than the second preset water level, an audible and visual alarm is issued and the predicted drainage volume is increased.
10. The drainage system according to claim 1, characterized in that, Also includes: Water quality testing module; The water quality detection module is used to monitor the pH value of the mine water in the mine water tank in real time. If the pH value is less than a first preset value, an alkaline substance is added to the mine water to increase the pH value. The alkaline substance includes sodium carbonate and sodium hydroxide.