Method for improving power supply reliability of underground coal mine equipment

Abstract

The invention provides a method for improving the power supply reliability of underground coal mine equipment, and belongs to the technical field of power supplies, and the method comprises the steps: inputting the operation data of a first power supply into a random forest model, and obtaining the operation state of the first power supply; a decision tree in the random forest model is divided from high to low according to specified fault levels; if the operation state of the first power supply is a fault state, switching the underground equipment power supply from the first power supply to a second power supply; inputting the operation data of the first power supply into a deep learning model to obtain a fault positioning result of the first power supply; the fault positioning result is used for indicating troubleshooting of the first power supply. According to the method for improving the power supply reliability of the underground coal mine equipment, the power supply reliability of the mining power supply can be improved.

Description

Technical Field

[0001] The present application belongs to the field of power supply technology, and more specifically, relates to a method for improving the power supply reliability of equipment in underground coal mines. Background Art

[0002] The production process of coal mining enterprises is closely linked. Power outages can lead to equipment downtime and process interruptions, causing waste of raw materials, product scrapping, and even safety accidents (such as coal mine gas accumulation and the risk of explosion in petrochemical equipment).

[0003] The power quick-cut device can quickly cut off the faulty power supply when a high-voltage power supply fails, quickly close the backup power supply or bus tie, and realize rapid switching of the power supply to ensure that the load can run continuously without disturbance or interruption.

[0004] In order to further improve the safe operation and power supply reliability of underground equipment, it is necessary to further improve the fault response speed of the power supply quick cut-off device. Summary of the Invention

[0005] The purpose of this application is to provide a method for improving the power supply reliability of equipment in underground coal mines, so as to improve the power supply reliability of mine power supply.

[0006] A first aspect of an embodiment of the present application provides a method for improving power supply reliability of equipment in a coal mine, comprising: Inputting the operating data of the first power supply into a random forest model to obtain an operating status of the first power supply; splitting the decision tree in the random forest model in order from high to low according to the specified fault level; If the operating status of the first power supply is a faulty state, the power supply of the downhole equipment is switched from the first power supply to the second power supply; the operating data of the first power supply is input into the deep learning model to obtain the fault location result of the first power supply; the fault location result is used to indicate troubleshooting of the first power supply.

[0007] A second aspect of an embodiment of the present application provides an apparatus for improving power supply reliability of equipment in underground coal mines, comprising: a state classification module, configured to input the operating data of the first power supply into a random forest model to obtain the operating state of the first power supply; wherein the decision tree in the random forest model is split in order from high to low according to the specified fault level; A power switching module is used to switch the power supply of the downhole equipment from the first power supply to the second power supply when the operating state of the first power supply is a fault state; the operating data of the first power supply is input into the deep learning model to obtain the fault location result of the first power supply; the fault location result is used to indicate the troubleshooting of the first power supply.

[0008] In a third aspect of an embodiment of the present application, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, the steps of the above-mentioned method for improving the power supply reliability of underground equipment in coal mines are implemented.

[0009] In a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned method for improving the power supply reliability of underground equipment in a coal mine are implemented.

[0010] The beneficial effects of a method for improving the power supply reliability of equipment in underground coal mines provided by an embodiment of the present application are: This embodiment of the application first uses a random forest model to perform an initial fault diagnosis. If a fault is detected in the first power source, the system promptly switches the power supply to the underground equipment to the second power source, ensuring reliable power supply for the equipment. A deep learning model is then used to locate the fault. This two-stage "coarse screening + fine judgment" approach combines the advantages of random forest and deep learning models, improving both fault response speed and fault location accuracy, thereby enhancing the reliability of mine power supply. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0012] Figure 1 A flowchart of a method for improving power supply reliability of equipment in underground coal mines provided in one embodiment of the present application; Figure 2 This is a structural block diagram of a device for improving the power supply reliability of equipment in underground coal mines provided in one embodiment of the present application; Figure 3 A schematic block diagram of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0013] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0014] In order to make the purpose, technical solutions and advantages of this application clearer, specific embodiments will be described below with reference to the accompanying drawings.

[0015] Please refer to Figure 1 , Figure 1 This is a flow chart of a method for improving the power supply reliability of equipment in underground coal mines, provided in one embodiment of the present application. The method can be performed by an electronic device. Specifically, the method may include: S101: Inputting the operating data of the first power supply into a random forest model to obtain the operating status of the first power supply; the decision tree in the random forest model is split in sequence from high to low according to the specified fault level.

[0016] In this embodiment, the operating data of the first power supply may include input voltage, output voltage, phase voltages, neutral point voltage, voltage fluctuation, voltage harmonics, total current, phase current, zero-sequence current, current rate of change, current harmonics, temperatures of key components (such as transformer windings, IGBT modules, and capacitors), ambient temperature, cooling system temperature, and vibration data (such as transformer vibration and mechanical component vibration). By inputting the operating data of the first power supply into a random forest model, the operating status of the first power supply can be quickly classified (e.g., normal or faulty). When a fault occurs in the first power supply, the random forest model can quickly identify the fault status of the first power supply.

[0017] In this embodiment, the specified fault levels may include level one faults, level two faults, and level three faults, among which level one faults are faults that directly threaten safety or cause permanent damage to the equipment, such as short circuit, overload, and other faults of the first power supply; level two faults are faults that affect equipment performance but can be operated for a short time, such as abnormal harmonic content, frequency deviation, and other faults of the first power supply; level three faults are faults that require timely maintenance due to component performance degradation or environmental abnormalities, such as decreased insulation resistance, increased vibration amplitude, and other faults.

[0018] In this embodiment, the decision tree is split in order from high to low according to the specified fault level, and the detection of the first-level fault can be given priority. When a first-level fault is detected, subsequent detection can be stopped immediately and the fault status can be output quickly to avoid the time delay caused by the detection of second-level and third-level faults.

[0019] S102: If the operating state of the first power supply is a fault state, switch the power supply of the downhole equipment from the first power supply to the second power supply; input the operating data of the first power supply into the deep learning model to obtain the fault location result of the first power supply; the fault location result is used to indicate the fault troubleshooting of the first power supply.

[0020] In this embodiment, when the operating state of the first power supply is determined to be a fault state according to the output of the random forest model, the power supply of the downhole equipment can be immediately switched to the second power supply to ensure reliable power supply of the equipment.

[0021] Specifically, the electronic device in this embodiment can be a controller of a power quick-cut device, which also includes a switching switch (such as a circuit breaker, a contactor, a solid-state relay, etc.). When the controller determines that the operating status of the first power supply is a fault state based on the output of the random forest model, the switching switch can be controlled to quickly cut off the first power supply and quickly close the second power supply.

[0022] Subsequently, the operating data of the first power supply (including operating data before and after the fault occurs) can be input into the deep learning model, and the feature mining capability of the deep learning model can be used to accurately locate the fault. The fault location results can guide maintenance personnel to quickly troubleshoot the problem and reduce maintenance time and costs.

[0023] As can be seen from the above, this embodiment first uses a random forest model to perform an initial fault diagnosis. If a fault is detected in the first power source, the underground equipment is promptly switched to the second power source, ensuring reliable power supply to the equipment. A deep learning model is then used to locate the fault. This two-stage "coarse screening + fine judgment" approach combines the advantages of random forest and deep learning models, improving both fault response speed and fault location accuracy, thereby enhancing the reliability of mine power supply.

[0024] In one embodiment of the present application, the deep learning model includes an input layer, a convolutional neural network, a long short-term memory network, a self-attention module, and an output layer connected in sequence. The operating data of the first power supply is input into the deep learning model to obtain a fault location result of the first power supply, including: The operating data of the first power supply is connected to the convolutional neural network after passing through the input layer to obtain local features; Input the local features into the long short-term memory network to obtain the time series features corresponding to multiple time steps; Input the temporal features into the attention module, calculate the weight of each time step, and perform weighted calculation on the temporal features corresponding to multiple time steps based on the weight to obtain the context vector; The context vector is output through the output layer to obtain the fault location result of the first power supply.

[0025] In this embodiment, the deep learning model adopts the architecture of convolutional neural network (CNN) + long short-term memory network (LSTM) + attention module, wherein the convolutional neural network can effectively capture the local features of the operating data of the first power supply (such as voltage spikes and current mutations), and is suitable for processing high-frequency anomalies in electrical signals; the long short-term memory network can model timing dependencies and capture fault evolution trends (such as a slow rise in temperature indicating component aging); the attention module calculates the weights of timing features, automatically focuses on key fault periods (such as a sudden change in current at the moment of short circuit), suppresses noise interference, and the context vector obtained by weighted calculation of timing features corresponding to multiple time steps integrates important information from different time steps, which is conducive to improving fault location accuracy.

[0026] Among them, the fault location results may include fault component information, such as rectifier module fault, filter capacitor fault, inverter fault, etc.; at the same time, the fault location results may also include fault type information, such as short circuit fault, overload fault, overheating fault, component aging, poor contact, etc.; in addition, the fault location results may also include fault severity assessment, such as level 1 fault, level 2 fault and level 3 fault, etc. The assessment results can be compared with the fault classification results output by the random forest model to verify the output results of the random forest model.

[0027] From the above, it can be concluded that this embodiment adopts the combined architecture of CNN+LSTM+attention module, which can effectively extract multi-scale fault features and focus on key time periods, which is conducive to improving fault location accuracy.

[0028] In one embodiment of the present application, the temporal features are input into the attention module, and the weight of each time step is calculated, including: Calculate the feature change rate corresponding to each time step respectively; the feature change rate corresponding to any time step is the change rate between the time series feature in any time step and the corresponding time series feature in the previous time step; The weight of each time step is calculated based on the feature change rate corresponding to each time step.

[0029] In this embodiment, taking any t-th time step as an example, the corresponding time series feature Corresponding features in the previous time step (t-1) The rate of change between can be expressed as: ; in, represents the rate of change of the feature corresponding to the t-th time step, The larger the value, the more dramatic the characteristic change at the tth time step, the greater its contribution to fault location, and the greater the weight of that time step. For example, if a short circuit occurs at the tth time step, the rate of change of the current characteristic is much higher than during normal operation. The weight automatically increases, guiding the model to focus on the data at the tth time step. For another example, for a poor contact fault, the intermittent contact causes the characteristic change rate to fluctuate periodically, and the weight distribution will emphasize the time step with the most significant fluctuation.

[0030] This embodiment calculates the weight of each time step based on the feature change rate corresponding to each time step, which can directly utilize the physical meaning of the feature change, avoid the misfocus of attention driven by pure data, and is conducive to improving the accuracy of fault location.

[0031] In one embodiment of the present application, the weight of each time step is calculated based on the feature change rate corresponding to each time step, including: For each time step: Get the reference value of the criticality of the time step; If there are multiple target features in the time step, and the correlation between the multiple target features meets the preset relationship, the reference value of the criticality of the time step is increased based on the first step length to obtain the criticality of the time step; the target feature is the feature whose corresponding feature change rate is greater than the change rate threshold; If at least one target feature exists in the time step and the correlation between the multiple target features does not meet the preset relationship, the reference value of the criticality of the time step is increased based on the second step length to obtain the criticality of the time step; the second step length is smaller than the first step length; Calculate the criticality of each time step based on the feature change rate corresponding to each time step; The weight of each time step is calculated based on its criticality.

[0032] In this embodiment, when calculating the weight of each time step, the criticality of each time step can be first calculated based on the feature change rate corresponding to each time step, and then the criticality of each time step is normalized to obtain the weight of each time step. The greater the feature change rate corresponding to each time step, the greater the corresponding criticality and weight. The formula for normalization calculation can be expressed as: ; in, represents the weight of the t-th time step, represents the criticality of the t-th time step, T represents the total number of time steps, and those skilled in the art can set the specific value of T according to actual needs.

[0033] The specific calculation process of the criticality of each time step is introduced below.

[0034] First, each time step can be set to have the same reference value (initial value) for criticality. Based on this, if multiple features undergo a sudden change within a certain time step, that is, the feature change rate corresponding to multiple features (target features) is greater than the change rate threshold, and the correlation between the multiple target features conforms to a preset relationship, reflecting the coupling relationship between the multiple features when a specific fault occurs, then the probability of the specific fault occurring is high. In this case, the reference value of the criticality for that time step can be increased by a larger first step length to serve as the criticality for that time step. The change rate threshold corresponding to each feature is a preset constant, and those skilled in the art can determine the specific value of the change rate threshold based on the historical operating data of the first power supply.

[0035] For example, when a short circuit fault occurs, the voltage drops sharply and the current rises sharply. If it is detected that in a certain time step, the characteristic change rates of the voltage and current are large, and the correlation between the voltage and current conforms to the preset relationship of a sharp drop in voltage and a sharp rise in current, it indicates that the probability of a short circuit fault is high. At this time, the reference value of the criticality of the time step can be increased by the first step length to highlight the importance of the time step.

[0036] Correspondingly, if at least one feature mutates within a certain time step, that is, the feature change rate corresponding to at least one feature (target feature) is greater than the change rate threshold, and when there are multiple target features, the correlation between the multiple target features does not meet the preset relationship (there is no coupling relationship), at this time, the reference value of the criticality of the time step can be increased by a smaller second step length as the criticality of the time step.

[0037] For example, in a certain time step, the characteristic change rates corresponding to the voltage and the electromagnetic vibration frequency of the transformer are both large, and there is no direct relationship between the voltage anomaly and the electromagnetic vibration frequency of the transformer. It is sufficient to pay a little attention to the characteristic changes in this period. At this time, the reference value of the criticality of this time step can be increased by the first step length.

[0038] The first step length and the second step length are both preset constants, and those skilled in the art can set specific values ​​of the first step length and the second step length according to actual needs.

[0039] From the above, it can be concluded that this embodiment calculates the weight of the attention module based on the feature change rate and the (physical) correlation between features, making the final deep learning model closer to the actual fault evolution logic, which can significantly improve the robustness and explainability of fault location.

[0040] In one embodiment of the present application, the input layer of the deep learning model includes multiple channels, and the operating data of the first power supply is connected to the convolutional neural network after passing through the input layer, including: After the electrical operating data, temperature operating data and environmental operating data in the operating data of the first power supply are respectively input into multiple channels of the input layer of the deep learning model, they are connected to the convolutional neural network.

[0041] In this embodiment, the electrical operating data within the first power supply's operating data includes real-time electrical parameters such as voltage, current, and frequency. The temperature operating data within the first power supply's operating data includes slowly changing parameters such as the temperature of the transformer and IGBT module and the temperature of the cooling system. The environmental operating data within the first power supply's operating data includes external environmental parameters such as humidity and vibration. Given the significant differences in the physical properties of different data types, this embodiment provides independent channels for each type of data. The data from each channel is processed through different convolutional layers for feature extraction. For example, for electrical operating data, features such as voltage spikes and current harmonics can be extracted. For temperature operating data, features such as the rate of temperature rise and thermal distribution anomalies can be extracted. For environmental operating data, features such as vibration frequency anomalies and humidity spikes can be extracted. Deeper convolution operations can be used to capture potential correlations between data by fusing features from different channels.

[0042] From the above, it can be concluded that this embodiment, through the multi-channel input layer design, can realize parallel processing and collaborative analysis of electrical operating data, temperature operating data and environmental operating data in the operating data of the first power supply, which helps to improve the accuracy and robustness of fault detection.

[0043] In one embodiment of the present application, the fault state includes a primary fault state, a secondary fault state, and a tertiary fault state. If the operating state of the first power supply is a fault state, switching the power supply of the downhole equipment from the first power supply to the second power supply includes: If the operating state of the first power supply is a level one fault state, switching the power supply of the downhole equipment from the first power supply to the second power supply within a first time period; If the operating state of the first power supply is a level 2 fault state, switching the power supply of the downhole equipment from the first power supply to the second power supply within a second time period; If the operating state of the first power supply is a level three fault state, switching the power supply of the downhole equipment from the first power supply to the second power supply within a third time period; The first duration, the second duration, and the third duration increase sequentially.

[0044] In this embodiment, considering that if power is switched immediately upon any fault, short-term disturbances (such as motor startup shock) may cause erroneous switching, this embodiment sets different switching times based on the fault level. For level 1 faults, which directly threaten equipment safety and require millisecond-level switching, the first switching time can be set to <50ms. For level 2 faults, retesting can be performed within a short period of time to avoid erroneous operation, and the second switching time can be set to 50-500ms. For level 3 faults, the switching can be completed within a longer time window, reducing unnecessary power disturbances.

[0045] From the above, it can be concluded that this embodiment determines the corresponding switching time based on different fault levels. On the premise of ensuring the power supply safety of underground equipment, it avoids frequent switching of power due to minor abnormalities, reduces interference with continuous production processes (such as coal mining machine operation and ventilation system), and reduces downtime losses.

[0046] In one embodiment of the present application, the operating data of the first power supply includes voltage data of a plurality of first monitoring points, current data of a plurality of second monitoring points, and temperature data of a plurality of third monitoring points; The method for improving the reliability of power supply to equipment in underground coal mines further includes: If the output current of the first power supply is less than or equal to a current threshold, setting the splitting weights of the voltage data of the plurality of first monitoring points to be greater than the splitting weights of the current data of the plurality of second monitoring points, and setting the splitting weights of the current data of the plurality of second monitoring points to be greater than the splitting weights of the temperature data of the plurality of third monitoring points, to obtain an adjusted random forest model; If the output current of the first power supply is greater than the current threshold, setting the splitting weights of the current data of the plurality of second monitoring points to be greater than the splitting weights of the temperature data of the plurality of third monitoring points, and setting the splitting weights of the temperature data of the plurality of third monitoring points to be greater than the splitting weights of the voltage data of the plurality of first monitoring points, to obtain an adjusted random forest model; Inputting the operating data of the first power supply into the random forest model to obtain the operating status of the first power supply, including: The operating data of the first power supply is input into the adjusted random forest model to obtain the operating status of the first power supply.

[0047] In this embodiment, voltage sensors can be installed at multiple first monitoring points to obtain multiple voltage data sets, including input voltage, output voltage, phase voltage, and neutral point voltage. Simultaneously, current sensors can be installed at multiple second monitoring points to obtain multiple current data sets, including total current, phase current, and zero-sequence current. Furthermore, temperature sensors can be installed at multiple third monitoring points to obtain multiple temperature data sets, including transformer winding temperature, IGBT module temperature, capacitor temperature, ambient temperature, and cooling system temperature. These voltage, current, and temperature data constitute the core data of the first power supply operation data.

[0048] Based on this, a current threshold (for example, 20% of the rated current) can be pre-set. When the output current of the first power supply is less than or equal to the current threshold, it indicates that the first power supply is in a light-load condition. In this condition, the load current is low, the line voltage drop is reduced, and the terminal voltage of the first power supply may be close to the no-load voltage. At this time, the impact of voltage fluctuations on the load equipment is more significant. In this case, when determining the operating status of the first power supply based on the random forest model, the split weight of the voltage data can be set to the maximum. In this way, when the decision tree splits, the voltage feature is more likely to be used as the split attribute of the root node or upper-level node, and is preferentially involved in fault diagnosis, thereby quickly identifying voltage anomalies (such as undervoltage and overvoltage).

[0049] Correspondingly, when the output current of the first power supply exceeds the current threshold, current and temperature become the dominant factors, significantly increasing the risk of overcurrent. Especially under heavy loads, high current can exacerbate conductor heating. In this case, when determining the operating status of the first power supply based on the random forest model, the split weight for current data can be set to the highest, followed by temperature data. This way, when the decision tree splits, current and temperature characteristics are more likely to serve as split attributes for the root node or upper-level nodes, giving them priority in fault diagnosis and enabling timely detection of faults such as overloads and short circuits.

[0050] From the above, it can be concluded that this embodiment dynamically adjusts the feature splitting weight based on the current threshold, so that the random forest model can adaptively adjust the splitting order of the decision tree according to the power load status, which meets the actual needs of power supply for underground equipment in coal mines and is conducive to improving the accuracy and efficiency of fault detection.

[0051] Corresponding to the above embodiment, a method for improving the power supply reliability of equipment in underground coal mines is provided. Figure 2 This is a structural block diagram of a device for improving the power supply reliability of equipment in coal mines provided in one embodiment of the present application. For ease of illustration, only the parts related to the embodiment of the present application are shown. Figure 2 The device 20 for improving the power supply reliability of equipment in underground coal mines includes: a state classification module 21 and a power switching module 22. The state classification module 21 is configured to input the operating data of the first power supply into a random forest model to obtain the operating state of the first power supply; the decision tree in the random forest model is split in order from high to low according to the specified fault level; The power switching module 22 is used to switch the power supply of the downhole equipment from the first power supply to the second power supply when the operating state of the first power supply is a fault state; the operating data of the first power supply is input into the deep learning model to obtain the fault location result of the first power supply; the fault location result is used to indicate the fault troubleshooting of the first power supply.

[0052] In one embodiment of the present application, the deep learning model includes an input layer, a convolutional neural network, a long short-term memory network, a self-attention module, and an output layer connected in sequence, and the power switching module 22 is specifically used to: The operating data of the first power supply is connected to the convolutional neural network after passing through the input layer to obtain local features; Input the local features into the long short-term memory network to obtain the time series features corresponding to multiple time steps; Input the temporal features into the attention module, calculate the weight of each time step, and perform weighted calculation on the temporal features corresponding to multiple time steps based on the weight to obtain the context vector; The context vector is output through the output layer to obtain the fault location result of the first power supply.

[0053] In one embodiment of the present application, the power switching module 22 is further configured to: Calculate the feature change rate corresponding to each time step respectively; the feature change rate corresponding to any time step is the change rate between the time series feature in any time step and the corresponding time series feature in the previous time step; The weight of each time step is calculated based on the feature change rate corresponding to each time step.

[0054] In one embodiment of the present application, the power switching module 22 is further configured to: For each time step: Get the reference value of the criticality of the time step; If there are multiple target features in the time step, and the correlation between the multiple target features meets the preset relationship, the reference value of the criticality of the time step is increased based on the first step length to obtain the criticality of the time step; the target feature is the feature whose corresponding feature change rate is greater than the change rate threshold; If at least one target feature exists in the time step and the correlation between the multiple target features does not meet the preset relationship, the reference value of the criticality of the time step is increased based on the second step length to obtain the criticality of the time step; the second step length is smaller than the first step length; Calculate the criticality of each time step based on the feature change rate corresponding to each time step; The weight of each time step is calculated based on its criticality.

[0055] In one embodiment of the present application, the input layer of the deep learning model includes multiple channels, and the power switching module 22 is further configured to: After the electrical operating data, temperature operating data and environmental operating data in the operating data of the first power supply are respectively input into multiple channels of the input layer of the deep learning model, they are connected to the convolutional neural network.

[0056] In one embodiment of the present application, the fault state includes a first-level fault state, a second-level fault state, and a third-level fault state. The power switching module 22 is further configured to: If the operating state of the first power supply is a level one fault state, switching the power supply of the downhole equipment from the first power supply to the second power supply within a first time period; If the operating state of the first power supply is a level 2 fault state, switching the power supply of the downhole equipment from the first power supply to the second power supply within a second time period; If the operating state of the first power supply is a level three fault state, switching the power supply of the downhole equipment from the first power supply to the second power supply within a third time period; The first duration, the second duration, and the third duration increase sequentially.

[0057] In one embodiment of the present application, the operating data of the first power supply includes voltage data of multiple first monitoring points, current data of multiple second monitoring points, and temperature data of multiple third monitoring points. The state classification module 21 is specifically configured to: If the output current of the first power supply is less than or equal to a current threshold, setting the splitting weights of the voltage data of the plurality of first monitoring points to be greater than the splitting weights of the current data of the plurality of second monitoring points, and setting the splitting weights of the current data of the plurality of second monitoring points to be greater than the splitting weights of the temperature data of the plurality of third monitoring points, to obtain an adjusted random forest model; If the output current of the first power supply is greater than the current threshold, setting the splitting weights of the current data of the plurality of second monitoring points to be greater than the splitting weights of the temperature data of the plurality of third monitoring points, and setting the splitting weights of the temperature data of the plurality of third monitoring points to be greater than the splitting weights of the voltage data of the plurality of first monitoring points, to obtain an adjusted random forest model; Inputting the operating data of the first power supply into the random forest model to obtain the operating status of the first power supply, including: The operating data of the first power supply is input into the adjusted random forest model to obtain the operating status of the first power supply.

[0058] See also Figure 3 , Figure 3This is a schematic block diagram of an electronic device provided in one embodiment of the present application. Figure 3 The electronic device 300 in the embodiment shown may include: one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memory 304 is used to store computer programs, which include program instructions. The processor 301 is used to execute the program instructions stored in the memory 304. The processor 301 is configured to call the program instructions to execute the functions of the modules / units in the above-mentioned device embodiments, such as Figure 2 The functions of the status classification module 21 and the power switching module 22 are shown.

[0059] It should be understood that in the embodiment of the present application, the processor 301 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc.

[0060] The input device 302 may include a touchpad, a fingerprint collection sensor (for collecting user fingerprint information and fingerprint direction information), a microphone, etc. The output device 303 may include a display (LCD, etc.), a speaker, etc.

[0061] The memory 304 may include a read-only memory and a random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include a non-volatile random access memory. For example, the memory 304 may also store specific values ​​of preset constants such as the rate of change threshold, first step size, second step size, and current threshold corresponding to each feature.

[0062] In a specific implementation, the processor 301, input device 302, and output device 303 described in the embodiment of the present application can execute the implementation method described in an embodiment of a method for improving the power supply reliability of underground equipment in coal mines provided in the embodiment of the present application, and can also execute the implementation method of the electronic device described in the embodiment of the present application, which will not be repeated here.

[0063] In another embodiment of the present application, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program. The computer program includes program instructions. When the program instructions are executed by a processor, all or part of the process of the method in the above embodiment is implemented. The computer program can also be used to instruct related hardware to complete the process. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of each of the above method embodiments are implemented. The computer program includes computer program code, which can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium.

[0064] The computer-readable storage medium can be an internal storage unit of the electronic device in any of the aforementioned embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. Furthermore, the computer-readable storage medium can include both an internal storage unit of the electronic device and an external storage device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium can also be used to temporarily store data that has been output or is about to be output.

[0065] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0066] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the electronic devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0067] In the several embodiments provided in this application, it should be understood that the disclosed electronic devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces or units, or can be an electrical, mechanical or other form of connection.

[0068] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0069] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0070] The above are only specific embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and such modifications or substitutions should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A method for improving the power supply reliability of equipment in underground coal mines, characterized in that: include: Inputting the operating data of the first power supply into the random forest model to obtain the operating status of the first power supply; The decision trees in the random forest model are split in order from high to low according to the specified fault levels; If the operating state of the first power supply is a fault state, switching the power supply of the downhole equipment from the first power supply to the second power supply; inputting the operating data of the first power supply into the deep learning model to obtain the fault location result of the first power supply; The fault location result is used to instruct to perform fault troubleshooting on the first power supply.

2. A method for improving power supply reliability of equipment in underground coal mines according to claim 1, characterized in that: The deep learning model includes an input layer, a convolutional neural network, a long short-term memory network, a self-attention module, and an output layer connected in sequence. Inputting the operating data of the first power supply into the deep learning model to obtain a fault location result of the first power supply includes: Connecting the operating data of the first power supply to a convolutional neural network after passing through the input layer to obtain local features; Inputting the local features into a long short-term memory network to obtain time series features corresponding to multiple time steps; Input the temporal features into the attention module, calculate the weight of each time step, and perform weighted calculation on the temporal features corresponding to multiple time steps based on the weight to obtain a context vector; The context vector is outputted through the output layer to obtain a fault location result of the first power supply.

3. A method for improving power supply reliability of equipment in underground coal mines according to claim 2, characterized in that: Inputting the temporal features into the attention module and calculating the weight of each time step includes: Calculate the feature change rate corresponding to each time step respectively; the feature change rate corresponding to any time step is the change rate between the time series feature in any time step and the corresponding time series feature in the previous time step; The weight of each time step is calculated based on the feature change rate corresponding to each time step.

4. A method for improving power supply reliability of equipment in coal mines according to claim 3, characterized in that: The weight of each time step is calculated based on the feature change rate corresponding to each time step, including: For each time step: Get the reference value of the criticality of the time step; If there are multiple target features in the time step, and the correlation between the multiple target features meets the preset relationship, the reference value of the criticality of the time step is increased based on the first step length to obtain the criticality of the time step; the target feature is a feature whose corresponding feature change rate is greater than the change rate threshold; If at least one target feature exists within the time step and the correlation between the multiple target features does not meet the preset relationship, the reference value of the criticality of the time step is increased based on the second step length to obtain the criticality of the time step; the second step length is smaller than the first step length; Calculate the criticality of each time step based on the feature change rate corresponding to each time step; The weight of each time step is calculated based on its criticality.

5. The method for improving power supply reliability of equipment in underground coal mines according to claim 2, characterized in that: The input layer of the deep learning model includes multiple channels, and the operation data of the first power supply is connected to the convolutional neural network after passing through the input layer, including: After the electrical operating data, temperature operating data and environmental operating data in the operating data of the first power supply are respectively input into multiple channels of the input layer of the deep learning model, they are connected to the convolutional neural network.

6. The method for improving power supply reliability of equipment in underground coal mines according to claim 1, characterized in that: The fault state includes a first-level fault state, a second-level fault state, and a third-level fault state. If the operating state of the first power supply is a fault state, switching the power supply of the downhole equipment from the first power supply to the second power supply includes: If the operating state of the first power supply is a level one fault state, switching the power supply of the downhole equipment from the first power supply to the second power supply within a first time period; If the operating state of the first power supply is a level 2 fault state, switching the power supply of the downhole equipment from the first power supply to the second power supply within a second time period; If the operating state of the first power supply is a level three fault state, switching the power supply of the downhole equipment from the first power supply to the second power supply within a third time period; The first duration, the second duration, and the third duration increase sequentially.

7. The method for improving power supply reliability of equipment in underground coal mines according to claim 1, characterized in that: The operation data of the first power supply includes voltage data of a plurality of first monitoring points, current data of a plurality of second monitoring points, and temperature data of a plurality of third monitoring points; The method for improving power supply reliability of equipment in underground coal mines further includes: If the output current of the first power supply is less than or equal to a current threshold, setting the splitting weights of the voltage data of the plurality of first monitoring points to be greater than the splitting weights of the current data of the plurality of second monitoring points, and setting the splitting weights of the current data of the plurality of second monitoring points to be greater than the splitting weights of the temperature data of the plurality of third monitoring points, to obtain an adjusted random forest model; If the output current of the first power supply is greater than a current threshold, setting the splitting weights of the current data of the plurality of second monitoring points to be greater than the splitting weights of the temperature data of the plurality of third monitoring points, and setting the splitting weights of the temperature data of the plurality of third monitoring points to be greater than the splitting weights of the voltage data of the plurality of first monitoring points, to obtain an adjusted random forest model; Inputting the operating data of the first power supply into the random forest model to obtain the operating state of the first power supply includes: The operating data of the first power supply is input into the adjusted random forest model to obtain the operating status of the first power supply.

8. A device for improving the power supply reliability of equipment in underground coal mines, characterized in that: include: a state classification module, configured to input the operating data of the first power supply into a random forest model to obtain an operating state of the first power supply; The decision trees in the random forest model are split in order from high to low according to the specified fault levels; A power switching module, configured to switch the power supply of the downhole equipment from the first power supply to the second power supply when the operating state of the first power supply is a fault state; Inputting the operating data of the first power supply into a deep learning model to obtain a fault location result of the first power supply; The fault location result is used to instruct to perform fault troubleshooting on the first power supply.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images