Intelligent working face remote monitoring method and system based on coal machine equipment

Abstract

The invention discloses an intelligent working face remote monitoring method and system based on coal machine equipment, and the method comprises the steps: collecting the operation data and environment data of the coal machine equipment, and carrying out the preprocessing; and analyzing the preprocessed data according to a multi-level hybrid integrated deep neural network algorithm to obtain a current risk coefficient. Different measures are taken for different risk conditions, and one-time monitoring is completed. According to the intelligent working face remote monitoring method and system based on the coal machine equipment provided by the invention, spatial and temporal features are extracted by using a convolutional layer in a deep learning model, a recursive layer captures a time sequence dependency relationship, and different weights are given to different features by an attention mechanism, so that the operation state and potential risks of the equipment are accurately evaluated. According to the step, the fusion analysis of high-dimensional data is realized through the multi-level design of the deep neural network model, and the accuracy and real-time performance of risk assessment are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial automation, and in particular to an intelligent working face remote monitoring method and system based on coal mining equipment. Background Art

[0002] With the rapid development of the coal mining industry, coal mining equipment is increasingly used in modern coal mining operations, and its operational stability and safety are crucial to coal mine production. Traditional coal mining equipment monitoring methods mainly rely on on-site manual inspections and single sensor data monitoring. This method is not only inefficient, but also difficult to capture equipment anomalies in complex operating environments in real time. In addition, with the increase in the depth and scale of coal mining, the diversity and harshness of the working face environment have put forward higher requirements for equipment monitoring. In recent years, with the development of Industrial Internet of Things (IIoT) and Artificial Intelligence (AI), remote monitoring and intelligent diagnosis technologies have gradually become important research directions in the field of coal mining equipment monitoring. Data acquisition, analysis and processing methods based on sensor networks are becoming the core technologies for realizing equipment status evaluation and fault prediction.

[0003] Although the existing intelligent monitoring systems have achieved data collection and analysis to a certain extent, there are still obvious shortcomings. First, the existing systems are often based only on a single data source or simple algorithms, and cannot fully reflect the complexity of the interaction between the equipment's operating status and the environment. Secondly, in terms of data processing, existing technologies mainly rely on simple rules or shallow machine learning models, which are difficult to handle multi-dimensional and highly complex data features, resulting in insufficient risk assessment accuracy. In addition, due to the lack of a dynamic response mechanism, the existing monitoring systems' responses to abnormal situations are mostly static presets, which makes it difficult to achieve accurate and flexible handling of different risk levels. These problems limit the further application and promotion of coal machinery equipment monitoring technology in complex working face environments. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is that the existing system usually relies on static rules to respond, and it is difficult to adjust the strategy according to the dynamic changes of risk levels. The present invention realizes accurate handling of different risk situations through a hierarchical response mechanism based on risk coefficients, thereby improving the security and adaptability of the system.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: an intelligent working face remote monitoring method based on coal machinery equipment, comprising: collecting coal machinery equipment operation data and environmental data, and performing preprocessing.

[0007] The preprocessed data is analyzed according to the multi-level hybrid integrated deep neural network algorithm to obtain the current risk coefficient.

[0008] Take different measures for different risk situations to complete a monitoring.

[0009] As a preferred solution of the intelligent working face remote monitoring method based on coal machinery equipment described in the present invention, the collection of coal machinery equipment operation data and environmental data includes collecting data through sensors, and the coal machinery equipment operation data includes speed, torque, temperature, vibration, current and voltage.

[0010] Environmental data includes air temperature, humidity, gas concentration, air pressure and illumination.

[0011] As a preferred solution of the intelligent working face remote monitoring method based on coal machinery equipment described in the present invention, the preprocessing includes data cleaning, data interpolation, normalization and feature extraction.

[0012] Data cleaning includes noise removal and outlier detection and elimination. Noise removal uses filtering algorithms to remove random noise in the data, and outlier detection and elimination uses statistical methods to monitor and eliminate abnormal data points.

[0013] Interpolation methods are used to complete missing data caused by sensor failure or communication interruption.

[0014] Normalization uses the minimum and maximum normalization method to map the original data to the range of 0 to 1, expressed as:

[0015]

[0016] Among them, x norm represents the normalized data value, x represents the original data value, and x min Indicates the minimum value of the data indicator, x max Indicates the maximum value of the data indicator.

[0017] Feature extraction is performed on the preprocessed data to extract key features that can represent the equipment operating status and environmental changes, including time domain features, frequency domain features and statistical features.

[0018] The variance and mean of the time domain features are calculated.

[0019] The frequency domain features are extracted by fast Fourier transform, which is expressed as:

[0020]

[0021] Among them, X(k0) represents the complex spectrum value of the kth frequency component, k represents the frequency index, and the value range is 0 to N-1. N represents the total number of sampling points of the signal, x(n0) represents the signal value at the nth time point, e -j2πkn / Nrepresents the complex exponential basis function, which is used to transform the time domain signal into the frequency domain. j represents the imaginary unit.

[0022] The statistical features are the maximum value, minimum value, and peak value of each data.

[0023] As a preferred solution of the intelligent working face remote monitoring method based on coal machinery equipment described in the present invention, wherein: the preprocessed data is analyzed according to the multi-level hybrid integrated deep neural network algorithm to obtain the current risk coefficient, and the multi-level hybrid integrated deep neural network algorithm includes an input layer, a convolutional layer, a recursive layer, an attention mechanism layer and an output layer.

[0024] As a preferred solution of the intelligent remote monitoring method for coal mining equipment in the present invention, the multi-level hybrid integrated deep neural network algorithm includes: n ] is input to the input layer of the multi-level hybrid integrated deep neural network, where f i represents the i-th feature.

[0025] The convolutional layer applies a 3D convolution operation to extract spatiotemporal features:

[0026]

[0027] Among them, C1 represents the output feature map of the first hidden layer, σ represents the activation function, and the ReLU function is selected. A, B, and C represent the sizes of the convolution kernel in three dimensions. W1(a, b, c0 represents the convolution kernel weight, which corresponds to the index of the three dimensions respectively. F(a+b, b+c, c) represents the local area value of the input feature in the convolution operation.

[0028] The recurrent layer uses LSTM units to process the output of the convolutional layer to capture the long-term dependencies in the time series, which can be expressed as:

[0029] H t =tanh(W f ·C 1,t +U f ·H t-1 +b f )⊙σ(W i ·C 1,t +U i ·H t-1 +b i )

[0030] Among them, H t represents the hidden state of the second hidden layer at time point t, tanh represents the hyperbolic tangent activation function, W f , W i Indicates the current input C 1,tThe weight matrix U f , U i Represents the previous hidden state H t-1 The weight matrix of b. f 、b i Represents the bias vector. ⊙ represents the element-wise multiplication operation. σ represents the sigmoid activation function, which is used for the gating mechanism.

[0031] The attention mechanism layer assigns different weights to different features, expressed as:

[0032]

[0033] Among them, A t represents the value of the attention weight vector at time point t, exp represents the exponential function, which is used to calculate the normalization of the weight. K represents the number of attention heads, α k represents the weight coefficient of the kth attention head, represents the kth feature of the second hidden layer output.

[0034] The output of the attention mechanism layer passes through the fully connected layer and applies the Softmax function to obtain the final risk coefficient:

[0035]

[0036] Among them, R represents the final calculated risk coefficient, Z represents the normalization factor, and the value range of the risk coefficient is guaranteed to be between 0 and 1. T represents the total number of time steps, A t represents the value of the attention weight at time point t, J represents the number of neurons in the fully connected layer, and β j represents the weight coefficient of the fully connected layer, φ j (· represents the activation function, and the Softplus function is selected.

[0037] As a preferred solution of the intelligent working face remote monitoring method based on coal machinery equipment described in the present invention, wherein: the current risk coefficient is obtained including that the value range of the risk coefficient R is [0,1].

[0038] When 0≤R<0.3, it is low risk.

[0039] When 0.3≤R<0.5, it is medium-low risk.

[0040] When 0.5≤R<0.7, it is medium-high risk.

[0041] When 0.7≤R≤1, it is high risk.

[0042] As a preferred solution of the intelligent remote monitoring method for coal mining equipment in the present invention, the following is described: different measures are taken for different risk situations, and one monitoring is completed, and in a low-risk situation, the system should perform operations including continuing to monitor the equipment operation data and environmental data in real time to ensure that the risk factor is maintained at a low level. The current data and risk factor are recorded in the database for subsequent analysis and trend prediction. Regular operation reports are generated to feedback the equipment operation status and potential optimization suggestions.

[0043] In low- to medium-risk situations, the system should send early warning notifications to maintenance personnel and management personnel via SMS, email, or mobile applications to alert them of potential risks. It is recommended to conduct routine inspections and maintenance of the equipment to ensure that all parts of the equipment are in good condition to prevent failures. According to specific risk factors, the operating parameters of the equipment can be appropriately adjusted, such as reducing the speed or torque, to reduce potential risks.

[0044] In medium- and high-risk situations, the system should perform operations such as issuing emergency warnings to all relevant personnel through multiple communication channels to ensure that information is quickly conveyed. Activate the automatic adjustment mechanism to adjust the equipment operating parameters to a safe range in real time, such as automatically reducing load or speeding up to prevent equipment overload or overheating. Arrange technicians to immediately conduct detailed inspections on site to check for potential faults and suspend equipment operation for maintenance if necessary.

[0045] In high-risk situations, the system should automatically start the emergency shutdown procedure of the equipment and immediately stop the equipment from running to prevent further deterioration of the risk. Send emergency alarm signals to all relevant personnel through sound and light alarm systems, mobile device push, etc., and start the emergency response procedure. Start the safety isolation system to cut off the connection between the equipment and the power supply or other key systems to ensure that the equipment stops operating completely. Record all relevant data of this high-risk event in detail for subsequent analysis and cause finding to prevent similar incidents from happening again. Start the predetermined emergency response plan and mobilize the emergency team to handle the scene to ensure the safety of personnel and equipment.

[0046] An intelligent working face remote monitoring system based on coal mining equipment, characterized in that:

[0047] The preprocessing module collects coal mining equipment operation data and environmental data for preprocessing.

[0048] The calculation module analyzes the preprocessed data according to the multi-level hybrid integrated deep neural network algorithm to obtain the current risk coefficient.

[0049] Take measures module to take different measures for different risk situations and complete a monitoring.

[0050] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0051] A computer-readable storage medium stores a computer program, which implements the steps of the method described above when executed by a processor.

[0052] The beneficial effects of the present invention are as follows: the convolutional layer in the deep learning model is used to extract spatiotemporal features, the recursive layer captures time series dependencies, and the attention mechanism assigns different weights to different features, thereby accurately assessing the operating status and potential risks of the equipment. This step realizes the fusion analysis of high-dimensional data through the multi-level design of the deep neural network model, significantly improving the accuracy and real-time performance of risk assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0054] Figure 1 An overall flow chart of an intelligent working face remote monitoring method and system based on coal mining machinery equipment is provided for the first embodiment of the present invention. DETAILED DESCRIPTION

[0055] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0056] Example 1, reference Figure 1 , which is an embodiment of the present invention, provides an intelligent working face remote monitoring method based on coal mining equipment, comprising:

[0057] S1: Collect coal mining equipment operation data and environmental data and perform preprocessing.

[0058] Data is collected through sensors, and the operating data of coal machinery equipment includes speed, torque, temperature, vibration, current and voltage.

[0059] Environmental data includes air temperature, humidity, gas concentration, air pressure and illumination.

[0060] It should be noted that the data collection is as follows:

[0061] Speed ​​(RPM): The speed of the coal machine is monitored in real time by a rotary encoder installed on the main shaft of the coal machine.

[0062] Torque (Nm): Use a torque sensor to measure the torque changes of the coal machine during operation.

[0063] Temperature (℃): Install temperature sensors at key locations (such as motor bearings, gearboxes, etc.) to monitor equipment temperature in real time.

[0064] Vibration(m / s 2 ): Use acceleration sensors to collect vibration signals during the operation of coal mining machines to evaluate the health status of the equipment.

[0065] Current and voltage (A and V): The working status of the coal machine electrical system is monitored by current sensors and voltage sensors.

[0066] Air temperature (°C): Use an ambient temperature sensor to monitor the air temperature around the work surface.

[0067] Humidity (%RH): Get the ambient humidity data through the humidity sensor.

[0068] Gas concentration (ppm): Install gas detection instruments to monitor the concentration of harmful gases (such as carbon monoxide, methane, etc.) in real time.

[0069] Air pressure (kPa): Use an air pressure sensor to measure changes in air pressure on the working surface.

[0070] Illuminance (lux): The light intensity of the work surface is obtained through the light sensor.

[0071] Preprocessing includes data cleaning, data interpolation, normalization and feature extraction.

[0072] Data cleaning includes noise removal and outlier detection and elimination. Noise removal uses filtering algorithms to remove random noise in the data, and outlier detection and elimination uses statistical methods to monitor and eliminate abnormal data points.

[0073] It should be noted that in the filtering process, the window size determines the balance between filtering effect and real-time performance. The window size needs to be reasonably selected based on the data sampling rate and the operating characteristics of the device. A larger window can remove noise more effectively, but may cause signal delay. A smaller window can maintain the real-time performance of the signal, but may not completely remove noise.

[0074] Furthermore, the 3σ principle is used for outlier detection. Assuming that the data follows a normal distribution, the data points within the 3σ range cover 99.7% of the data, and those outside this range are considered outliers. In specific applications, it is necessary to ensure that the data is close to a normal distribution, otherwise it may be necessary to adjust the threshold or use other anomaly detection methods.

[0075] Interpolation methods are used to complete missing data caused by sensor failure or communication interruption.

[0076] Minimum-maximum normalization is applicable to the case where the data distribution range is known and there are no outliers. The original data is mapped to the range of 0 to 1, expressed as:

[0077]

[0078] Among them, x norm represents the normalized data value, x represents the original data value, and x min Indicates the minimum value of the data indicator, x max Indicates the maximum value of the data indicator.

[0079] Feature extraction is performed on the preprocessed data to extract key features that can represent the equipment operating status and environmental changes, including time domain features, frequency domain features and statistical features.

[0080] The variance and mean of the time domain features are calculated.

[0081] The frequency domain features are extracted by fast Fourier transform, which is expressed as:

[0082]

[0083] Among them, X(k0) represents the complex spectrum value of the kth frequency component, k represents the frequency index, and the value range is 0 to N-1. N represents the total number of sampling points of the signal, x(n0) represents the signal value at the nth time point, and e -j2πkn / N represents the complex exponential basis function, which is used to transform the time domain signal into the frequency domain. j represents the imaginary unit.

[0084] The statistical features are the maximum value, minimum value, and peak value of each data.

[0085] S2: Analyze the preprocessed data according to the multi-level hybrid integrated deep neural network algorithm to obtain the current risk coefficient.

[0086] The multi-level hybrid integrated deep neural network algorithm includes input layer, convolution layer, recursive layer, attention mechanism layer and output layer.

[0087] The preprocessed feature vector F = [f1, f2, ..., f n] is input to the input layer of the multi-level hybrid integrated deep neural network, where f i represents the i-th feature.

[0088] The convolutional layer applies a 3D convolution operation to extract spatiotemporal features:

[0089]

[0090] Among them, C1 represents the output feature map of the first hidden layer, σ represents the activation function, and the ReLU function is selected. A, B, and C represent the sizes of the convolution kernel in three dimensions. W1(a, b, c) represents the convolution kernel weight, which corresponds to the index of the three dimensions respectively. F(a+b, b+c, c) represents the local area value of the input feature in the convolution operation.

[0091] The recurrent layer uses LSTM units to process the output of the convolutional layer to capture the long-term dependencies in the time series, which can be expressed as:

[0092] H t =tanh(W f ·C 1,t +U f ·H t-1 +b f )⊙σ(W i ·C 1,t +U i ·H t-1 +b i )

[0093] Among them, H t represents the hidden state of the second hidden layer at time point t, tanh represents the hyperbolic tangent activation function, W f , W i Indicates the current input C 1,t The weight matrix U f , U i Represents the previous hidden state H t-1 The weight matrix of b. f 、b i Represents the bias vector. ⊙ represents the element-wise multiplication operation. σ represents the sigmoid activation function, which is used for the gating mechanism.

[0094] The attention mechanism layer assigns different weights to different features, expressed as:

[0095]

[0096] Among them, A t represents the value of the attention weight vector at time point t, exp represents the exponential function, which is used to calculate the normalization of the weight. K represents the number of attention heads, α k represents the weight coefficient of the kth attention head, represents the kth feature of the second hidden layer output.

[0097] The output of the attention mechanism layer passes through the fully connected layer and applies the Softmax function to obtain the final risk coefficient:

[0098]

[0099] Among them, R represents the final calculated risk coefficient, Z represents the normalization factor, and the value range of the risk coefficient is guaranteed to be between 0 and 1. T represents the total number of time steps, A t represents the value of the attention weight at time point t, J represents the number of neurons in the fully connected layer, and β j represents the weight coefficient of the fully connected layer, φ j (·) represents the activation function, and the Softplus function is selected.

[0100] The value range of the risk coefficient R is [0,1].

[0101] When 0≤R<0.3, it is low risk.

[0102] When 0.3≤R<0.5, it is medium-low risk.

[0103] When 0.5≤R<0.7, it is medium-high risk.

[0104] When 0.7≤R≤1, it is high risk.

[0105] S3: Take different measures for different risk situations and complete a monitoring.

[0106] Different measures are taken for different risk situations. To complete a monitoring, in low-risk situations, the system should continue to monitor the equipment operation data and environmental data in real time to ensure that the risk factor remains at a low level. The current data and risk factor are recorded in the database for subsequent analysis and trend prediction. Regular operation reports are generated to provide feedback on equipment operation status and potential optimization suggestions.

[0107] In low- to medium-risk situations, the system should send early warning notifications to maintenance personnel and management personnel via SMS, email, or mobile applications to alert them of potential risks. It is recommended to conduct routine inspections and maintenance of the equipment to ensure that all parts of the equipment are in good condition to prevent failures. According to specific risk factors, the operating parameters of the equipment can be appropriately adjusted, such as reducing the speed or torque, to reduce potential risks.

[0108] In medium- and high-risk situations, the system should perform operations such as issuing emergency warnings to all relevant personnel through multiple communication channels to ensure that information is quickly conveyed. Activate the automatic adjustment mechanism to adjust the equipment operating parameters to a safe range in real time, such as automatically reducing load or speeding up to prevent equipment overload or overheating. Arrange technicians to immediately conduct detailed inspections on site to check for potential faults and suspend equipment operation for maintenance if necessary.

[0109] In high-risk situations, the system should automatically start the emergency shutdown procedure of the equipment and immediately stop the equipment from running to prevent further deterioration of the risk. Send emergency alarm signals to all relevant personnel through sound and light alarm systems, mobile device push, etc., and start the emergency response procedure. Start the safety isolation system to cut off the connection between the equipment and the power supply or other key systems to ensure that the equipment stops operating completely. Record all relevant data of this high-risk event in detail for subsequent analysis and cause finding to prevent similar incidents from happening again. Start the predetermined emergency response plan and mobilize the emergency team to handle the scene to ensure the safety of personnel and equipment.

[0110] The above embodiments also include an intelligent working face remote monitoring system based on coal mining equipment, specifically:

[0111] The preprocessing module collects coal mining equipment operation data and environmental data for preprocessing.

[0112] The calculation module analyzes the preprocessed data according to the multi-level hybrid integrated deep neural network algorithm to obtain the current risk coefficient.

[0113] Take measures module to take different measures for different risk situations and complete a monitoring.

[0114] The computer device may be a server. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data cluster data of the power monitoring system. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a remote monitoring method for an intelligent working face based on coal machinery equipment is realized.

[0115] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0116] Example 2 is an embodiment of the present invention, which provides an intelligent working face remote monitoring method and system based on coal machinery equipment. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through simulation experiments.

[0117] This embodiment aims to verify the superiority of the intelligent working face remote monitoring method based on coal mining equipment. The experiment selected the key parameters of a coal mine working face as the experimental object, and compared the existing traditional monitoring technology with the intelligent remote monitoring system of the present invention. The experimental data was collected by equipment sensors in the actual operating environment to reflect the actual working conditions.

[0118] Test preparation:

[0119] The experiment selected the operating parameters (speed, temperature, vibration) and environmental data (gas concentration, humidity) of the coal machine equipment as the core indicators for monitoring. Sensor types include high-precision rotary encoders, temperature sensors, three-axis vibration sensors, gas concentration detectors, and humidity sensors, all of which are connected to the data acquisition system. The experimental conditions simulate the normal mining environment of a coal mine, in which three faults and two extreme environmental changes are randomly added to verify the system's anti-interference ability and real-time response performance.

[0120] Existing technology implementation process:

[0121] A single sensor is used to directly collect the above key parameters, and an independent data storage module is used to record various data.

[0122] Data analysis is processed using a simple rule engine, such as an alarm if a vibration parameter exceeds a preset threshold.

[0123] There is no dynamic response to equipment abnormalities, only manual inspection and processing after the alarm.

[0124] Implementation process of the present invention:

[0125] Data collection and preprocessing: Equipment and environmental parameters are collected through high-precision sensors. Data preprocessing includes noise filtering (Kalman filtering), outlier removal (3σ principle) and interpolation completion (spline interpolation method), and finally a normalized feature vector is generated.

[0126] Risk coefficient assessment: The feature vector is input into a multi-level hybrid integrated deep neural network model, the spatiotemporal features are extracted through the convolutional layer, the long-term dependencies are captured through the recursive layer, the feature weights are assigned by the attention mechanism, and the risk coefficient is finally output.

[0127] Grading response: According to the risk factor, it is divided into low risk, medium-low risk, medium-high risk and high risk levels. The system dynamically adjusts the operating parameters of the coal machine equipment according to the level, implements shutdown protection when necessary, and sends an early warning to the remote control center. The experimental results are shown in Table 1.

[0128] Table 1 Experimental results

[0129]

[0130]

[0131] From the table data, it can be seen that both the prior art and the present invention can cover basic operating parameters and environmental data in data collection. However, through further analysis, it can be found that the present invention is significantly superior to the prior art in data processing accuracy, response capability and dynamic adjustment capability.

[0132] In the existing technology, the vibration and gas concentration data fluctuate greatly during the sampling process. For example, the vibration values ​​of sampling 3 and sampling 4 increased significantly (6.5m / s 2 Up to 7.0m / s 2 ), failed to capture the changing trend in time.

[0133] The present invention uses data cleaning and interpolation to complete the process, so that the vibration data is smoother and closer to the actual working conditions (for example, the vibration value of sampling 2 is only 5.8m / s 2 , 0.2m / s less than the existing technology 2 ), the fluctuation of gas concentration data is reduced, and the stability of the data is enhanced.

[0134] The existing technology only alarms for over-threshold events and cannot reflect the trend of equipment status changes. When the gas concentration from sampling 3 to sampling 4 increased from 240ppm to 260ppm, the system failed to respond in advance.

[0135] The present invention uses a multi-level hybrid integrated deep neural network to analyze the full amount of data, and the risk assessment results are more accurate. For example, when sampling 3 (gas concentration 230ppm), the equipment operating parameters have been adjusted to prevent further deterioration.

[0136] The present invention dynamically adjusts the coal machine speed (495RPM to 485RPM) during sampling 2 and sampling 3, effectively controlling the vibration amplitude of the equipment (5.8m / s 2 Up to 6.2m / s 2 ), while reducing the temperature rise trend (62°C). The existing technology does not have this adjustment function, vibration and temperature continue to rise, and equipment risks increase.

[0137] It can be seen from the table that the present invention has a better correction effect on abnormal data of humidity and gas concentration. For example, in the fourth sampling stage, the humidity recorded by the prior art is 100%, while the present invention corrects it to 97%, which is more in line with reality.

[0138] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An intelligent working face remote monitoring method based on coal mining equipment, characterized in that: include: Collect coal mining equipment operation data and environmental data and perform pre-processing; Analyze the preprocessed data according to the multi-level hybrid integrated deep neural network algorithm to obtain the current risk factor; Take different measures for different risk situations to complete a monitoring.

2. The intelligent working face remote monitoring method based on coal mining equipment according to claim 1 is characterized in that: The collecting of coal mining equipment operation data and environmental data includes collecting data through sensors, and the coal mining equipment operation data includes speed, torque, temperature, vibration, current and voltage; Environmental data includes air temperature, humidity, gas concentration, air pressure and illumination.

3. The intelligent working face remote monitoring method based on coal mining equipment according to claim 2 is characterized in that: The preprocessing includes data cleaning, data interpolation, normalization and feature extraction; Data cleaning includes noise removal and outlier detection and elimination. Noise removal uses filtering algorithms to remove random noise in the data, and outlier detection and elimination uses statistical methods to monitor and eliminate abnormal data points. For missing data caused by sensor failure or communication interruption, interpolation methods are used to complete the missing data; Normalization uses the minimum and maximum normalization method to map the original data to the range of 0 to 1, expressed as: Among them, x norm represents the normalized data value, x represents the original data value, and x min Indicates the minimum value of the data indicator, x max Indicates the maximum value of the data indicator; Perform feature extraction on the preprocessed data to extract key features that can represent the equipment operating status and environmental changes, including time domain features, frequency domain features, and statistical features; Calculate variance and mean of time domain features; The frequency domain features are extracted by fast Fourier transform, which is expressed as: Where X(k) represents the complex spectrum value of the kth frequency component, k represents the frequency index, ranging from 0 to N-1; N represents the total number of sampling points of the signal, x(n) represents the signal value at the nth time point, and e -j2πkn / N represents the complex exponential basis function, which is used to convert the time domain signal to the frequency domain; j represents the imaginary unit; The statistical features are the maximum value, minimum value, and peak value of each data.

4. The intelligent working face remote monitoring method based on coal mining equipment according to claim 3 is characterized in that: The preprocessed data is analyzed according to the multi-level hybrid integrated deep neural network algorithm to obtain the current risk coefficient, which includes an input layer, a convolutional layer, a recursive layer, an attention mechanism layer and an output layer.

5. The intelligent working face remote monitoring method based on coal mining equipment according to claim 4 is characterized in that: The multi-level hybrid integrated deep neural network algorithm includes: n ] is input to the input layer of the multi-level hybrid integrated deep neural network, where f i represents the i-th feature; The convolutional layer applies a 3D convolution operation to extract spatiotemporal features: Among them, C1 represents the output feature map of the first hidden layer, σ represents the activation function, and the ReLU function is selected; A, B, and C represent the sizes of the convolution kernel in three dimensions; W1(a, b, c) represents the convolution kernel weight, which corresponds to the index of the three dimensions respectively; F(a+b, b+c, c0 represents the local area value of the input feature in the convolution operation; The recurrent layer uses LSTM units to process the output of the convolutional layer to capture the long-term dependencies in the time series, which can be expressed as: H t fishy ( W f ·C 1,t +U f ·H t-1 +b f) ⊙σ ( W i ·C 1,t +U i ·H t-1 +b i ) Among them, H t represents the hidden state of the second hidden layer at time point t, tanh represents the hyperbolic tangent activation function, W f , W i Indicates the current input C 1,t The weight matrix of f , U i Represents the previous hidden state H t-1 The weight matrix of f , b i represents the bias vector; ⊙ represents the element-by-element multiplication operation; σ represents the sigmoid activation function, which is used for the gating mechanism; The attention mechanism layer assigns different weights to different features, expressed as: Among them, A t represents the value of the attention weight vector at time point t, exp represents the exponential function, which is used to calculate the normalization of the weight; K represents the number of attention heads, α k represents the weight coefficient of the kth attention head, represents the kth feature of the second hidden layer output; The output of the attention mechanism layer passes through the fully connected layer and applies the Softmax function to obtain the final risk coefficient: Among them, R represents the final calculated risk coefficient, Z represents the normalization factor, which ensures that the value range of the risk coefficient is between 0 and 1; T represents the total number of time steps, and A t represents the value of the attention weight at time point t, J represents the number of neurons in the fully connected layer, and β j represents the weight coefficient of the fully connected layer, φ j ( · ) Represents the activation function, and the Softplus function is selected.

6. The intelligent working face remote monitoring method based on coal mining equipment according to claim 5 is characterized in that: The obtaining of the current risk coefficient includes that the value range of the risk coefficient R is [0,1]; When 0≤R<0.3, it is low risk; When 0.3≤R<0.5, it is medium-low risk; When 0.5≤R<0.7, it is medium-high risk; When 0.7≤R≤1, it is high risk.

7. The intelligent working face remote monitoring method based on coal mining equipment according to claim 6 is characterized in that: The said different measures are taken for different risk situations, and completing one monitoring includes, in the case of low risk, the operation that the system should perform includes, continuing to monitor the equipment operation data and environmental data in real time to ensure that the risk factor is maintained at a low level; Record current data and risk factors into the database for subsequent analysis and trend prediction; Generate regular operation reports to provide feedback on equipment operation status and potential optimization suggestions; In low- to medium-risk situations, the system should send early warning notifications to maintenance personnel and management personnel via SMS, email or mobile applications to remind them of potential risks; recommend routine inspection and maintenance of equipment to ensure that all parts of the equipment are in good condition to prevent failures; and appropriately adjust the operating parameters of the equipment according to specific risk factors, such as reducing the speed or torque, to reduce potential risks; In medium- and high-risk situations, the system should perform actions such as issuing emergency warnings to all relevant personnel through multiple communication channels to ensure that information is conveyed quickly; Start the automatic adjustment mechanism to adjust the equipment operating parameters to a safe range in real time, such as automatically reducing load or speeding up, to prevent equipment overload or overheating; arrange for technicians to immediately go to the site for detailed inspection, troubleshoot potential faults, and suspend equipment operation for maintenance if necessary; In high-risk situations, the system should perform the following operations: automatically start the equipment's emergency shutdown procedure to immediately stop the equipment from running to prevent further risk; send emergency alarm signals to all relevant personnel through sound and light alarm systems, mobile device push, etc., and start the emergency response procedure; start the safety isolation system to cut off the connection between the equipment and the power supply or other critical systems to ensure that the equipment completely stops operating; record all relevant data of the high-risk event in detail for subsequent analysis and cause finding to prevent similar incidents from happening again; start the predetermined emergency response plan, mobilize the emergency team for on-site processing, and ensure the safety of personnel and equipment.

8. An intelligent working face remote monitoring system based on coal mining equipment using the method according to any one of claims 1 to 7, characterized in that: The preprocessing module collects the coal machine equipment operation data and environmental data for preprocessing; The calculation module analyzes the preprocessed data according to the multi-level hybrid integrated deep neural network algorithm to obtain the current risk coefficient; Take measures module to take different measures for different risk situations and complete a monitoring.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: 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 having a computer program stored thereon, 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