Coal mine equipment fault prediction system and method based on big data

Abstract

The invention relates to the technical field of equipment fault prediction, in particular to a coal mine equipment fault prediction system and method based on big data, and the method comprises the following steps: collecting the thickness of a biological membrane on the surface of equipment, synchronously collecting the vibration acceleration of the equipment and the environmental hydrogen sulfide concentration, and constructing a corrosion multi-mode data set; performing electrochemical noise separation on the corrosion multi-modal data set, and constructing a corrosion activity fingerprint spectrum; decoupling a coupling effect of mechanical vibration and microbial corrosion, and generating a corrosion-fatigue damage correlation map; and constructing a dual-channel life prediction model based on the corrosion-fatigue damage correlation map, and setting an early warning mechanism. Microbial corrosion characteristic current is accurately extracted through wavelet packet decomposition and wavelet entropy analysis, a corrosion-vibration heterogeneous graph is established by adopting a graph attention network, corrosion-induced structural vibration enhancement effect analysis is realized, and compared with a traditional prediction method based on a single signal, the method has the advantages that the method is simple and convenient to operate. And the problems of limited local corrosion influence range, unstable vibration response and the like are effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment failure prediction, and in particular to a coal mine equipment failure prediction system and method based on big data. Background Art

[0002] During the operation of equipment, corrosion damage and fatigue cracks of metal structures are the two main factors affecting the safety and life of equipment, especially underground coal mine-related equipment. Traditional corrosion prediction methods are usually based on electrochemical measurement methods, relying on laboratory tests or regular on-site monitoring, and it is difficult to achieve real-time tracking of the corrosion process. On the other hand, fatigue damage prediction mainly uses stress analysis and crack growth models, such as the Paris formula, but this method assumes that crack growth is only affected by mechanical loads and does not consider the weakening effect of corrosion on the microstructure of the material, resulting in a decrease in prediction accuracy. In addition, corrosion and fatigue failure are often coupled with each other. Stress concentration caused by local corrosion will accelerate crack initiation, and structural vibration may also promote the diffusion of corrosive media, forming a vicious cycle. Therefore, it is difficult to accurately predict the comprehensive life of equipment by relying solely on a single failure mode analysis.

[0003] The existing crack growth model does not consider the influence of corrosion, resulting in deviations in fatigue life prediction. Classic models such as the Paris crack growth law and Miner linear cumulative damage theory mainly calculate the crack growth rate based on mechanical stress and cyclic loading, and do not fully consider the accelerating effect of corrosion on crack growth. Under actual working conditions, local corrosion can cause microscopic damage to the material and affect the stress distribution at the crack tip, making the fatigue crack growth rate much higher than the theoretical prediction. Summary of the invention

[0004] The present invention provides a coal mine equipment failure prediction system and method based on big data.

[0005] The coal mine equipment failure prediction method based on big data includes the following steps: S1: Deploy ultrasonic sensors on the metal surface of the equipment to collect the biofilm thickness on the equipment surface, and simultaneously collect the equipment vibration acceleration and environmental hydrogen sulfide concentration to construct a corrosion multimodal data set; S2: performing electrochemical noise separation on the corrosion multimodal data set, extracting characteristic current pulses of microbial corrosion by wavelet entropy algorithm, and constructing a corrosion activity fingerprint spectrum; S3: Input the corrosion activity fingerprint spectrum into the graph attention network to decouple the coupling effect of mechanical vibration and microbial corrosion, and generate a corrosion-fatigue damage association map; S4: A dual-channel life prediction model is constructed based on the corrosion-fatigue damage correlation map, and the composite failure trajectory curve including the chemical corrosion rate and the mechanical crack growth rate is output. An early warning mechanism is set based on the composite failure trajectory curve.

[0006] Optionally, the S1 specifically includes: S11: The ultrasonic sensor emits ultrasonic pulses, monitors reflected echoes, and calculates the sound wave propagation time difference: ,in, is the biofilm thickness, is the ultrasonic propagation velocity in the biofilm, which is 1550 m / s. is the propagation time delay of ultrasonic wave; S12: Install a three-axis MEMS vibration sensor on the surface of the device, and set the vibration sampling frequency to 10kHz; S13: Arrange multiple hydrogen sulfide electrochemical sensor arrays around the equipment to obtain environmental hydrogen sulfide concentration data; S14: unifying the sampling time base of each sensor through a hardware clock synchronizer to generate a raw data stream with a unified timestamp; S15: Generate a corrosion multimodal data set including biofilm thickness, equipment vibration acceleration, and ambient hydrogen sulfide concentration.

[0007] Optionally, the S2 specifically includes: S21: Based on the db4 wavelet basis function, the electrochemical noise signal in the corrosion multimodal data set is decomposed at multiple levels. In the wavelet decomposition, each layer of the signal is split into a low-frequency part and a high-frequency part to form multiple sub-bands. The decomposition process continues for eight levels, so that the signal is divided into multiple frequency ranges. S22: Calculate the wavelet entropy value of each sub-band: After obtaining the target sub-band signal, analyze the energy distribution characteristics, use wavelet entropy analysis to quantify the complexity of the signal, determine whether the signal includes corrosion characteristics, count the proportion of different energy components in each sub-band, obtain the energy distribution characteristics, calculate the wavelet entropy value based on the energy distribution characteristics, measure the randomness and structural information of the signal, the larger the wavelet entropy value, the higher the signal complexity, and it may contain more corrosion characteristics. In all sub-band signals, set the entropy value threshold. When the wavelet entropy value of a sub-band exceeds the entropy value threshold, it is determined that the current sub-band includes effective microbial corrosion information and is retained for subsequent feature extraction; S23: adaptively calculate the signal strength threshold of each sub-band to adapt it to signal changes in different environments. When the actual signal strength exceeds the signal strength threshold, it is determined to be a corrosion current pulse. Among all signal segments exceeding the signal strength threshold, key pulse parameters related to the corrosion process are extracted, including: Pulse amplitude, used to characterize the intensity of the corrosion reaction; Pulse duration, reflecting the length of time of the corrosion reaction; Pulse rise time, which is used to measure the suddenness of the corrosion process; S24: After extracting all key pulse parameters, a three-dimensional feature vector is constructed. The three-dimensional feature vector includes the average amplitude of the pulse, the total duration, and the change characteristics of the rise time. The corrosion activity is classified using the K-means clustering algorithm. The corrosion signal is divided into low activity level, medium activity level, and high activity level. The level division is based on the pulse amplitude and duration indicators. The clustering results are mapped to the time-frequency dimension to form a corrosion activity fingerprint spectrum. In the corrosion activity fingerprint spectrum, the horizontal axis represents time, the vertical axis represents characteristic frequency, and the color change reflects the different levels of corrosion activity.

[0008] In the calculation of the wavelet entropy value of each sub-band in S22, for each sub-band , its wavelet entropy value Calculated as: ,in, For the Tier Energy proportion of each node: ,in, Represents the energy of the node. When , the layer is judged to contain effective corrosion features. is the entropy threshold.

[0009] Optionally, the signal strength threshold Calculated as: ,in, For the The standard deviation of the layer signal, is the number of sampling points of this layer.

[0010] Optionally, the S3 specifically includes: S31: Construct a heterogeneous graph structure to describe the association between corrosion nodes and vibration nodes. The corrosion node is a pulse feature vector extracted from the corrosion activity fingerprint spectrum; the vibration node is a frequency band energy entropy vector extracted from the vibration time-frequency spectrum. After the nodes are constructed, define the association between corrosion and vibration, and use a weighting mechanism based on time difference to assign an edge weight to each corrosion-vibration connection edge. The value of the edge weight decays as the time difference between the corrosion node and the vibration node increases, ensuring that events that are close in time have higher correlation. S32: Construct a dual-channel graph attention network based on a heterogeneous graph structure, including a corrosion feature enhancement channel and a vibration-corrosion coupling channel, where: The corrosion feature enhancement channel is used to analyze the mutual correlation between corrosion nodes. The multi-head attention mechanism is used to calculate the attention of each corrosion node to other corrosion nodes, and the feature expression of the corrosion nodes is adjusted. Each corrosion node calculates its own correlation with other corrosion nodes, and uses the attention weighting method to update its own features to enhance the corrosion features. The vibration-corrosion coupling channel is used to analyze the impact of vibration signals on the corrosion process. The gating mechanism is used to screen effective associated edges. The gating mechanism calculates a dynamic weight based on the joint characteristics of corrosion and vibration to correct the initial time-related edge weight. The corrected edge weight is used to adjust the impact of vibration on the corrosion node. S33: Using singular value decomposition method to reduce the dimension of embedded features of corrosion-vibration nodes and decouple coupling effects, including calculating the mutual information between corrosion and vibration in the feature space after dimension reduction. Mutual information measures the statistical correlation between vibration signal and corrosion signal. The larger the value, the stronger the coupling between the two. When the mutual information exceeds the predetermined threshold, it indicates that the vibration signal and the corrosion signal have a strong dependence relationship, indicating that the corrosion process aggravates the fatigue vibration of the equipment. S43: After decoupling the relationship between corrosion and vibration, a corrosion-fatigue damage correlation map is constructed to describe the coupling mode between the two. The correlation map includes a horizontal axis and a vertical axis, where: The horizontal axis of the spectrum represents the corrosion activity intensity, which is derived from the chromaticity information of the corrosion activity fingerprint spectrum; The vertical axis of the spectrum represents the vibration energy entropy, which comes from the integration result of the time-frequency spectrum; The edge weight is defined as: ,in, and Represent the timestamps of corrosion pulse events and vibration events, respectively. is the time-dependent scaling factor.

[0011] The corrosion feature enhancement channel uses a multi-head attention mechanism (4 heads) to calculate the spatial correlation between corrosion nodes: ,in, and is the learnable parameter matrix, is the feature dimension, is the attention weight between the eroded nodes, indicating The corrosion node is The attention scores of the eroded nodes are used to calculate the spatial correlation between nodes. Represents a corrosion node The characteristic vector of Represents a corrosion node The eigenvector of The vibration-corrosion coupling channel screens valid correlation edges through the gated attention mechanism: ; ; in, is the Sigmoid function, is the gating weight matrix; represents the vector concatenation operation, is the gated attention value of the vibration-corrosion coupling channel, is the corrected edge weight, represents the pulse feature vector, Represents the frequency band energy entropy vector.

[0012] Optionally, the S4 specifically includes: S41: Construct a chemical corrosion rate prediction channel; extract the average amplitude of the pulse current from the corrosion activity fingerprint spectrum, obtain the hydrogen sulfide concentration gradient and the biofilm thickness growth rate, use the time convolution network to learn the dynamic characteristics of chemical corrosion, capture the corrosion change trend on different time scales, and establish a mapping relationship between the average amplitude, hydrogen sulfide concentration gradient and corrosion rate. Among them, the pulse current amplitude is used to measure the basic intensity of corrosion, the hydrogen sulfide concentration gradient affects the nonlinear change of the corrosion reaction rate, and the biofilm thickness growth rate is used to reflect the contribution of microbial activity to corrosion; S42: Construct a mechanical crack growth rate prediction channel; extract the vibration-corrosion mutual information from the corrosion-vibration correlation map, obtain the maximum principal stress and stress intensity factor amplitude to characterize the direct effect of mechanical load on crack growth. In the crack growth prediction, the traditional Paris law is used and the acceleration effect of corrosion on crack growth is considered; S43: Composite failure trajectory synthesis; establish a mapping relationship between the time scale of chemical corrosion and the number of mechanical cycles of mechanical fatigue, introduce time axis alignment processing, associate the evolution process of chemical corrosion with the expansion process of fatigue cracks, and analyze multiple failure modes in the same time coordinate system. This mapping relationship adjusts the accumulation method of fatigue cycles by considering the impact of corrosion on the operating frequency of the equipment. After the time axis alignment processing, calculate the composite failure trajectory curve, which describes the cumulative damage of corrosion and crack expansion during the service process. The equivalent service time calculation method is used to combine the chemical corrosion time and the number of mechanical cycles to define the composite damage degree; S44: Early warning mechanism; after obtaining the composite damage degree, the early warning mechanism is set according to different damage thresholds. When the damage degree exceeds the set damage threshold but still does not reach the critical value, the first-level early warning is triggered and planned maintenance is recommended to prevent further deterioration. If the damage degree continues to increase and its growth rate exceeds the critical value, the second-level early warning is triggered and emergency shutdown measures are taken.

[0013] Optionally, in the early warning mechanism of S44: when When the alarm is triggered, the first level warning is triggered; When the following two conditions are met, the second level warning is triggered: ,in, represents the composite damage degree, represents the composite damage degree change rate, It's hours, is the equivalent service time.

[0014] The coal mine equipment fault prediction system based on big data is used to implement the above-mentioned coal mine equipment fault prediction method based on big data, and includes the following modules: Data acquisition module: The biofilm thickness on the equipment surface is collected through ultrasonic sensors, and the equipment vibration acceleration and environmental hydrogen sulfide concentration are simultaneously acquired to construct a corrosion multimodal data set; Signal processing module: used to perform electrochemical noise separation on corrosion multimodal data sets, and use wavelet entropy algorithm to extract microbial corrosion characteristic current pulses to generate corrosion activity fingerprint spectrum; Coupling analysis module: used to input the corrosion activity fingerprint spectrum into the graph attention network, decouple the coupling effect of mechanical vibration and microbial corrosion, and generate the corrosion-fatigue damage association map; Life prediction and early warning module: used to build a dual-channel life prediction model based on the corrosion-fatigue damage correlation map, calculate the composite failure trajectory curve of chemical corrosion rate and mechanical crack growth rate, and set a dynamic threshold early warning mechanism based on the composite failure trajectory curve to achieve coal mine equipment failure prediction and risk early warning.

[0015] Compared with the prior art, the beneficial effects of the present invention include: The present invention constructs a corrosion-fatigue composite damage prediction model based on multimodal data such as biofilm thickness, corrosion current pulses, vibration time-frequency characteristics, and hydrogen sulfide concentration gradient. The characteristic current of microbial corrosion is accurately extracted through wavelet packet decomposition and wavelet entropy analysis, and a corrosion-vibration heterogeneous graph is established using a graph attention network to achieve analysis of the corrosion-induced structural vibration enhancement effect. Compared with traditional prediction methods based on a single signal, the present invention can jointly model in three dimensions of time, frequency, and space, effectively overcoming the problems of limited local corrosion influence range and unstable vibration response, thereby improving the prediction accuracy of corrosion-fatigue damage.

[0016] The present invention introduces the vibration-corrosion mutual information on the basis of the Paris crack growth formula, extracts the corrosion-induced features in the vibration signal through singular value decomposition, and calculates the dynamic correction of the crack growth rate by the corrosion influencing factor. This method can accurately identify the local stress concentration area caused by microbial corrosion and predict the nonlinear expansion trend of the crack, thereby reducing the prediction error of the crack initiation time and effectively extending the service life assessment accuracy of the equipment.

[0017] The present invention proposes a time axis alignment algorithm, which realizes the synchronous calculation of corrosion rate and crack propagation rate by establishing a nonlinear mapping relationship between chemical corrosion rate and fatigue cycle number, thereby ensuring the consistency of failure prediction. At the same time, a composite damage degree calculation model is used to quantify the cumulative damage degree of the equipment, and a dynamic threshold warning mechanism is introduced to trigger different levels of warnings when the damage degree exceeds the set value. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 A schematic diagram of a prediction method flow chart of an embodiment of the present invention; Figure 2 A schematic diagram of a dual-channel attention network according to an embodiment of the present invention; Figure 3 Schematic diagram of the prediction system functional modules according to an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0021] It should be noted that the references to "one embodiment", "embodiment", "exemplary embodiments", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0022] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0023] like Figure 1-Figure 2 As shown, the coal mine equipment failure prediction method based on big data includes the following steps: S1: Deploy ultrasonic sensors on the metal surface of the equipment to collect the biofilm thickness on the equipment surface, and simultaneously collect the equipment vibration acceleration and environmental hydrogen sulfide concentration to construct a corrosion multimodal data set; S2: Electrochemical noise separation is performed on the corrosion multimodal data set, and the characteristic current pulses of microbial corrosion are extracted by wavelet entropy algorithm to construct the corrosion activity fingerprint spectrum; S3: Input the corrosion activity fingerprint spectrum into the graph attention network to decouple the coupling effect of mechanical vibration and microbial corrosion, and generate a corrosion-fatigue damage association map; S4: A dual-channel life prediction model is constructed based on the corrosion-fatigue damage correlation map, and the composite failure trajectory curve including the chemical corrosion rate and the mechanical crack growth rate is output. An early warning mechanism is set based on the composite failure trajectory curve.

[0024] S1 specifically includes: S11: The ultrasonic sensor emits ultrasonic pulses, monitors the reflected echo, and calculates the sound wave propagation time difference: ,in, is the biofilm thickness, is the ultrasonic propagation velocity in the biofilm, which is 1550 m / s. is the propagation time delay of ultrasonic wave; S12: A three-axis MEMS vibration sensor is conformally installed on the surface of the device within 5 cm of the biofilm sensor, and the vibration sampling frequency is set to 10 kHz; S13: Arrange multiple hydrogen sulfide electrochemical sensor arrays around the device, with the electrochemical sensor arrays densely arranged at a spacing of 30 cm to obtain environmental hydrogen sulfide concentration data; S14: The sampling time base of each sensor is unified through the hardware clock synchronizer, the clock deviation is controlled within ±100 μs, and the original data stream with a unified timestamp is generated; S15: Generate a corrosion multimodal data set including biofilm thickness, equipment vibration acceleration, and ambient hydrogen sulfide concentration.

[0025] S2 specifically includes: S21: Based on the db4 wavelet basis function, the electrochemical noise signal in the corrosion multimodal data set is decomposed at multiple levels. In the wavelet decomposition, each layer of the signal is split into a low-frequency part and a high-frequency part to form multiple sub-bands. The decomposition process continues for eight levels, so that the signal is divided into multiple frequency ranges. In all sub-bands, it is necessary to focus on the characteristic signals related to microbial corrosion. Experimental studies have shown that the current pulses generated during microbial corrosion are mainly concentrated in a specific frequency range, so the sub-band signals from the third to the fourth layer are selected, and the corresponding frequency range is about 125 to 500 Hz. These sub-band signals contain electrochemical perturbations caused by microbial activity and can better reflect the characteristics of the microbial corrosion process; electrochemical noise is the current or potential fluctuation caused by the corrosion process. The electrochemical noise signal mainly comes from the electrochemical current fluctuation measured during the biofilm thickness monitoring process. Vibration acceleration and environmental hydrogen sulfide concentration are not electrochemical noise signals themselves, but will indirectly affect the corrosion process. Mechanical vibration will change the ion distribution of the corrosion interface, resulting in current fluctuations. High concentrations of hydrogen sulfide can accelerate the growth of biofilms, thereby enhancing electrochemical noise; S22: Calculate the wavelet entropy value of each sub-band. After obtaining the target sub-band signal, analyze the energy distribution characteristics. Use wavelet entropy analysis to quantify the complexity of the signal to determine whether the signal includes corrosion characteristics. Count the proportion of different energy components in each sub-band to obtain energy distribution characteristics. Calculate the wavelet entropy value based on the energy distribution characteristics to measure the randomness and structural information of the signal. The larger the wavelet entropy value, the higher the signal complexity, which may contain more corrosion characteristics. Set the entropy value threshold in all sub-band signals. When the wavelet entropy value of a sub-band exceeds the entropy value threshold, it is determined that the current sub-band includes effective microbial corrosion information and is retained for subsequent feature extraction. S23: For the screened sub-band signals, extract the microbial corrosion current pulse signal. Due to the large volatility of electrochemical noise signals, the traditional fixed threshold method is difficult to accurately extract weak corrosion pulses. Therefore, the dynamic threshold method is used for adaptive judgment. The signal strength threshold of each sub-band is adaptively calculated to adapt to the signal changes in different environments. When the actual signal strength exceeds the signal strength threshold, it is determined to be a corrosion current pulse. Among all signal segments that exceed the signal strength threshold, extract the key pulse parameters related to the corrosion process, including: Pulse amplitude, used to characterize the intensity of the corrosion reaction; Pulse duration, reflecting the length of time of the corrosion reaction; Pulse rise time, which is used to measure the suddenness of the corrosion process; These parameters can fully describe the characteristics of transient current pulses during microbial corrosion and provide basic data for subsequent analysis; S24: After extracting all key pulse parameters, a three-dimensional feature vector is constructed. The three-dimensional feature vector includes the average amplitude of the pulse, the total duration, and the change characteristics of the rise time. The K-means clustering algorithm is used to classify the corrosion activity, and the corrosion signal is divided into low activity level, medium activity level and high activity level. The level division is based on the pulse amplitude and duration indicators to ensure that the classification results can effectively correspond to microbial corrosion behaviors of different intensities. The clustering results are mapped to the time-frequency dimension to form a corrosion activity fingerprint spectrum. In the corrosion activity fingerprint spectrum, the horizontal axis represents time, the vertical axis represents characteristic frequency, and the color change reflects the different levels of corrosion activity. Through the fingerprint spectrum, the intensity changes of microbial corrosion at different time points can be clearly observed, providing a reliable basis for equipment failure prediction and maintenance.

[0026] The electrochemical noise signal in the corrosion multimodal data set is decomposed by wavelet packet, and the db4 wavelet basis function is selected for 8-layer decomposition. The characteristic subbands of microbial corrosion are extracted at the 3rd and 4th layers (corresponding to the frequency range of 125-500Hz); the db4 wavelet has 4 vanishing moments, which can better extract the mutation characteristics (corrosion current pulses). For the input signal, the wavelet packet decomposition uses a pair of filters, including a low-pass filter (approximate component) and a high-pass filter (detail component). Each decomposition will generate two sub-signals, including high-frequency components and low-frequency components. The signal is decomposed into 8 layers using the db4 wavelet basis. Each layer of decomposition will divide the signal into low frequency (approximate component) and high frequency (detail component). After 8 layers of decomposition, there are 256 sub-bands. The frequency range of each layer is recursively divided according to the binary method, and the characteristic subbands of microbial corrosion are extracted at the 3rd and 4th layers (corresponding to the frequency range of 125-500Hz). The 125-500Hz frequency band is selected as the characteristic interval of the microbial corrosion signal.

[0027] In calculating the wavelet entropy value of each subband in S22, for each subband , its wavelet entropy value Calculated as: ,in, For the Tier Energy proportion of each node: ,in, Represents the energy of the node. When , the layer is judged to contain effective corrosion features. is the entropy threshold.

[0028] Signal strength threshold Calculated as: ,in, For the The standard deviation of the layer signal, is the number of sampling points of this layer; For all signals exceeding the signal strength threshold The following key parameters are extracted from the pulse: Pulse amplitude : ; Duration : ; Rise time : ; in, and are the start and end time of the pulse, is the time point when the maximum pulse amplitude occurs, is a current signal, indicating time Instantaneous current value for detecting corrosion pulses: When constructing the corrosion activity fingerprint spectrum, it is necessary to count all pulse parameters and generate a three-dimensional feature vector : ,in, is the average amplitude of all pulses, Taking the logarithm of the total duration of all pulses, is the standard deviation of the rise time; K-means clustering is used to classify corrosion activity into three levels: low / medium / high: Low activity: ; Medium activity: ; Highly active: ; Map the clustering results to the time-frequency dimension to generate a fingerprint spectrum, including: Horizontal axis: time ; Vertical axis: characteristic frequency ; The color scale indicates the intensity of corrosion activity from blue (low) to red (high).

[0029] S3 specifically includes: S31: Construct a heterogeneous graph structure to describe the association between corrosion nodes and vibration nodes. The corrosion node is a pulse feature vector extracted from the corrosion activity fingerprint spectrum; the vibration node is a frequency band energy entropy vector extracted from the vibration time-frequency spectrum; after constructing the node, define the association between corrosion and vibration. Since corrosion and vibration events may have a certain offset in time, a weighting mechanism based on time difference is adopted to assign an edge weight to each corrosion-vibration connection edge. The value of the edge weight decays as the time difference between the corrosion node and the vibration node increases, ensuring that events close in time have a higher correlation; S32: Construct a dual-channel graph attention network based on a heterogeneous graph structure, including a corrosion feature enhancement channel and a vibration-corrosion coupling channel, where: The corrosion feature enhancement channel is used to analyze the mutual correlation between corrosion nodes. The multi-head attention mechanism is used to calculate the attention of each corrosion node to other corrosion nodes, and the feature expression of the corrosion nodes is adjusted. Each corrosion node calculates its own correlation with other corrosion nodes, and uses the attention weighting method to update its own features, enhance the corrosion features, and make them more consistent with the actual corrosion damage pattern. The vibration-corrosion coupling channel is used to analyze the impact of vibration signals on the corrosion process. Since vibration signals may contain a lot of irrelevant or noisy information, when calculating the corrosion-vibration correlation, the gating mechanism is first used to screen the effective correlation edges. The gating mechanism calculates a dynamic weight based on the joint characteristics of corrosion and vibration to correct the initial time-correlated edge weights, making the final corrosion-vibration coupling relationship more accurate. Finally, the corrected edge weights are used to adjust the degree of influence of vibration on the corrosion node; S33: Using singular value decomposition method to reduce the dimension of embedded features of corrosion-vibration nodes and decouple coupling effects, including calculating the mutual information between corrosion and vibration in the feature space after dimension reduction. Mutual information measures the statistical correlation between vibration signal and corrosion signal. The larger the value, the stronger the coupling between the two. When the mutual information exceeds the predetermined threshold, it indicates that the vibration signal and the corrosion signal have a strong dependence relationship, indicating that the corrosion process aggravates the fatigue vibration of the equipment. S43: After decoupling the relationship between corrosion and vibration, a corrosion-fatigue damage correlation map is constructed to describe the coupling mode between the two. The correlation map includes a horizontal axis and a vertical axis, where: The horizontal axis of the spectrum represents the corrosion activity intensity, which is derived from the chromaticity information of the corrosion activity fingerprint spectrum; The vertical axis of the spectrum represents the vibration energy entropy, which comes from the integration result of the time-frequency spectrum; The association strength encoding is visualized by color: In the strong coupling region, that is, the area with high corrosion-vibration mutual information, it is marked with red contour lines.

[0030] The independent action area, that is, the area where there is no obvious correlation between corrosion and vibration, is marked with a blue background.

[0031] In order to further identify the critical points that may cause equipment damage, the critical value of fatigue crack initiation is marked on the map. When the weighted combination of corrosion activity and vibration energy entropy in a specific frequency band exceeds the highest point, the equipment is considered to be at a high risk of fatigue crack initiation and maintenance or protective measures are required.

[0032] When building a heterogeneous graph structure, two types of nodes are defined: Corrosion node: Pulse feature vector extracted from the corrosion activity fingerprint spectrum: ; Vibration node: frequency band energy entropy vector extracted from the vibration time-frequency spectrum: , the energy entropy values ​​of the 3kHz, 5kHz, and 8kHz frequency bands in the vibration time-frequency spectrum are extracted separately to measure the complexity of the signal energy distribution in this frequency band; The edge weight is defined as: ; in, and Represent the timestamps of corrosion pulse events and vibration events, respectively. is the time-dependent scaling factor.

[0033] The corrosion feature enhancement channel uses a multi-head attention mechanism (4 heads) to calculate the spatial correlation between corrosion nodes: ,in, and is the learnable parameter matrix, is the feature dimension, , is the attention weight between the eroded nodes, indicating The corrosion node is The attention scores of the eroded nodes are used to calculate the spatial correlation between nodes. Represents a corrosion node The characteristic vector of Represents a corrosion node The eigenvector of The vibration-corrosion coupling channel screens valid correlation edges through the gated attention mechanism: ; ; in, is the Sigmoid function, is the gating weight matrix; represents the vector concatenation operation, is the gated attention value of the vibration-corrosion coupling channel, is the corrected edge weight, represents the pulse feature vector, Represents the frequency band energy entropy vector.

[0034] To decouple the coupling effects: perform singular value decomposition (SVD) on the node embeddings: ,in, For the The output feature matrix of the layer network only retains the first three singular vectors to construct the decoupled subspace. Represents the singular value decomposition (SVD) result: is the left singular vector matrix, representing the main characteristic directions of the corrosion and vibration nodes, is a singular value matrix, indicating the importance of each principal component. It is a right singular vector matrix, which represents the projection relationship between features at different nodes, dimensionality reduction and decoupling corrosion-vibration features after singular value decomposition, and the original embedded features It may contain coupled information of corrosion and vibration, but it may also have redundant or irrelevant parts. After decomposition, only the first three singular vectors are retained to construct the decoupled subspace, that is, to retain the main correlation features while reducing noise interference. The feature subspace after SVD dimensionality reduction can accurately represent the core relationship between corrosion and vibration without being affected by data redundancy.

[0035] Computing the vibration-corrosion mutual information in the decoupled subspace : ; when: hour, is a predetermined threshold value, judging that there is a strong coupling between vibration and corrosion, represents the joint probability distribution and the vibration node characteristics and corrosion node features The probability of co-occurrence, represents the marginal probability distribution of vibration node characteristics, Marginal probability distribution of corrosion node features.

[0036] In the corrosion-fatigue damage correlation diagram: 1. Horizontal axis: corrosion activity intensity (fingerprint spectrum chromaticity value); 2. Vertical axis: vibration energy entropy (integral value of time-frequency spectrum); 3. Association strength coding: Red contour: When , it is marked as a strong coupling region; Blue area: When , it is marked as an independent action area; 4. Mark the critical point of fatigue crack initiation: ,in, As an indicator of fatigue crack initiation under the combined effects of corrosion and vibration, when the value exceeds 0.75, the equipment is judged to have entered a high-risk fatigue damage stage. The 5kHz frequency band is usually related to the fatigue crack propagation vibration characteristics of the material. The energy changes of the high-frequency vibration signal within this frequency band can reflect the crack initiation and propagation process.

[0037] The frequency band energy entropy vector extracted from the vibration time-frequency spectrum is as follows: 1. Time-frequency spectrum generation: for the original vibration signal Perform a short-time Fourier transform: ; in, It is a Nuttall window with a window length of 50ms and a frequency resolution of 0.5Hz (which meets the requirements for high-frequency vibration characteristic detection of coal mining equipment). It represents the time-frequency spectrum, which is the representation of the vibration signal in the time-frequency domain obtained by short-time Fourier transform, and represents the energy distribution at time t and frequency f.

[0038] 2. Divide the key frequency bands based on the equipment failure mechanism: 3kHz: 2.8-3.2kHz: corresponds to the characteristic frequency of bearing outer ring defects; 5kHz: 4.8-5.2kHz: corresponds to the gear meshing subharmonic resonance frequency; 8kHz: 7.6-8.4kHz: corresponds to the high-frequency resonance frequency of microcracks.

[0039] 3. Frequency band energy calculate: Integrate the energy over each frequency band: ,in, Indicates the integration time, which refers to the time window length when calculating the frequency band energy. , covering the characteristic frequency fluctuation range, represents the center frequency, is the spectrum value after Fourier transform, Take 3kHZ, 5kHZ, 8kHZ, Indicates the window start time, Represent the integral differential elements of time variable and frequency variable respectively; 4. Energy entropy value solution: Normalization: , is the normalized band energy; Calculate the energy entropy of the frequency band: ; Final band energy entropy vector: .

[0040] S4 specifically includes: S41: Construct a chemical corrosion rate prediction channel; extract the average amplitude of the pulse current from the corrosion activity fingerprint spectrum, obtain the hydrogen sulfide concentration gradient and the biofilm thickness growth rate, use the time convolution network to learn the dynamic characteristics of chemical corrosion, capture the corrosion change trend on different time scales, and establish a mapping relationship between the average amplitude, hydrogen sulfide concentration gradient and corrosion rate. Among them, the pulse current amplitude is used to measure the basic intensity of corrosion, the hydrogen sulfide concentration gradient affects the nonlinear change of the corrosion reaction rate, and the biofilm thickness growth rate is used to reflect the contribution of microbial activity to corrosion; S42: Construct a mechanical crack growth rate prediction channel; extract the vibration-corrosion mutual information from the corrosion-vibration correlation map, obtain the maximum principal stress and stress intensity factor amplitude to characterize the direct effect of mechanical load on crack growth. In the crack growth prediction, based on the traditional Paris law and considering the acceleration effect of corrosion on crack growth, the crack growth rate depends not only on the amplitude of the stress intensity factor, but also on the vibration-corrosion mutual information. When the coupling degree of corrosion and vibration is high, the crack growth rate will increase significantly, thereby accelerating material failure; S43: Composite failure trajectory synthesis; establish a mapping relationship between the time scale of chemical corrosion and the number of mechanical cycles of mechanical fatigue, introduce time axis alignment processing, associate the evolution process of chemical corrosion with the expansion process of fatigue cracks, and analyze multiple failure modes in the same time coordinate system. This mapping relationship adjusts the accumulation method of fatigue cycles by considering the impact of corrosion on the operating frequency of the equipment. After the time axis alignment processing, calculate the composite failure trajectory curve, which describes the cumulative damage of corrosion and crack expansion during the service process. The equivalent service time calculation method is used to combine the chemical corrosion time and the number of mechanical cycles to define the composite damage degree. The composite damage degree takes into account the influence of corrosion rate and crack length. The damage degree is calculated using a nonlinear combination method to avoid the irrationality that may be caused by simple linear superposition. S44: Early warning mechanism; after obtaining the composite damage degree, the early warning mechanism is set according to different damage thresholds. When the damage degree exceeds the set damage threshold but still does not reach the critical value, the first-level early warning is triggered and planned maintenance is recommended to prevent further deterioration. If the damage degree continues to increase and its growth rate exceeds the critical value, the second-level early warning is triggered and emergency shutdown measures are taken to prevent sudden failure of the equipment.

[0041] The threshold setting of the early warning mechanism refers to the international material durability standards and the failure assessment standards of pipeline projects to ensure the reliability and applicability of the early warning system. The threshold of the first-level warning corresponds to the safe range of the equipment durability limit, while the threshold of the second-level warning corresponds to the critical point of failure in engineering applications. By dynamically adjusting the early warning threshold, false alarms and missed alarms can be effectively reduced and the prediction accuracy can be improved.

[0042] The input features when constructing the chemical corrosion rate prediction channel include: : Average pulse amplitude in the corrosion activity fingerprint spectrum; : Hydrogen sulfide concentration gradient, measuring the hydrogen sulfide concentration at different locations through a hydrogen sulfide electrochemical sensor array , and then calculate the rate of change of concentration along space: ,in, is the difference in hydrogen sulfide concentration measured by adjacent sensors, is the sensor spacing; : Biofilm thickness growth rate, is the unit change in biofilm thickness, is the time increment for a unit change in biofilm thickness; The calculation of chemical corrosion rate is based on the time convolution network TCN: , in, is the proportional coefficient (empirical parameter) in the corrosion rate calculation, which is used to adjust the influence of the average pulse amplitude on the chemical corrosion rate. is the normalization parameter of hydrogen sulfide concentration gradient, which is used to normalize the effect of hydrogen sulfide concentration gradient on corrosion rate. is the influence coefficient of the biofilm thickness growth rate, which indicates the contribution of biofilm thickness change to the chemical corrosion rate. , , , are material correlation coefficients.

[0043] The input features when constructing the mechanical crack growth rate prediction channel include: : Vibration-corrosion mutual information; : Maximum principal stress, measured by strain sensor, based on the material stress condition: ,in, is the external load, is the cross-sectional area under load; : The magnitude of the stress intensity factor, based on the crack size and load variation, follows the classical formula: ,in, is the maximum principal stress, is the crack length; Modified Paris's law model (coupled corrosion and crack growth effects): ,in, , , (represents the coupling effect coefficient), is the change in crack length, indicating the growth of the crack in one loading cycle. It represents a small increment in the number of loading cycles, representing the fatigue load cycle step experienced during crack growth.

[0044] Composite failure trajectory synthesis: Time axis alignment process: The chemical corrosion time scale is established by the following formula Number of mechanical cycles The mapping relationship between them is: ,in, Base frequency of the device , represents the corrosion-frequency coupling coefficient; Failure trajectory curve parameterization: Horizontal axis: Equivalent service time : ,in, (is the weight factor); Vertical axis: composite damage degree : ,in, (represents the critical corrosion rate), (represents the crack threshold length).

[0045] In the early warning mechanism: when When the alarm is triggered, the first level warning is triggered; When the following two conditions are met, the second level warning is triggered: ,in, represents the composite damage degree, represents the composite damage degree change rate, It's hours, is the equivalent service time.

[0046] like Figure 3 As shown, the coal mine equipment fault prediction system based on big data is used to implement the above prediction method, including the following modules: Data acquisition module: The biofilm thickness on the equipment surface is collected through ultrasonic sensors, and the equipment vibration acceleration and environmental hydrogen sulfide concentration are simultaneously acquired to construct a corrosion multimodal data set; Signal processing module: used to perform electrochemical noise separation on corrosion multimodal data sets, and use wavelet entropy algorithm to extract microbial corrosion characteristic current pulses to generate corrosion activity fingerprint spectrum; Coupling analysis module: used to input the corrosion activity fingerprint spectrum into the graph attention network, decouple the coupling effect of mechanical vibration and microbial corrosion, and generate the corrosion-fatigue damage association map; Life prediction and early warning module: used to build a dual-channel life prediction model based on the corrosion-fatigue damage correlation map, calculate the composite failure trajectory curve of chemical corrosion rate and mechanical crack growth rate, and set a dynamic threshold early warning mechanism based on the composite failure trajectory curve to achieve coal mine equipment failure prediction and risk early warning.

[0047] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0048] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A coal mine equipment failure prediction method based on big data, characterized in that: The following steps are involved: S1: Deploy ultrasonic sensors on the metal surface of the equipment to collect the biofilm thickness on the equipment surface, and simultaneously collect the equipment vibration acceleration and environmental hydrogen sulfide concentration to construct a corrosion multimodal data set; S2: performing electrochemical noise separation on the corrosion multimodal data set, extracting characteristic current pulses of microbial corrosion by wavelet entropy algorithm, and constructing a corrosion activity fingerprint spectrum; S3: Input the corrosion activity fingerprint spectrum into the graph attention network to decouple the coupling effect of mechanical vibration and microbial corrosion, and generate a corrosion-fatigue damage association map; S4: A dual-channel life prediction model is constructed based on the corrosion-fatigue damage correlation map, and the composite failure trajectory curve including the chemical corrosion rate and the mechanical crack growth rate is output. An early warning mechanism is set based on the composite failure trajectory curve.

2. The method for predicting coal mine equipment failure based on big data according to claim 1, characterized in that: The S1 specifically includes: S11: The ultrasonic sensor emits ultrasonic pulses, monitors reflected echoes, and calculates the sound wave propagation time difference: ,in, is the biofilm thickness, is the ultrasonic propagation velocity in the biofilm, which is 1550 m / s. is the propagation time delay of ultrasonic wave; S12: Install a three-axis MEMS vibration sensor on the surface of the device, and set the vibration sampling frequency to 10kHz; S13: Arrange multiple hydrogen sulfide electrochemical sensor arrays around the equipment to obtain environmental hydrogen sulfide concentration data; S14: unifying the sampling time base of each sensor through a hardware clock synchronizer to generate a raw data stream with a unified timestamp; S15: Generate a corrosion multimodal data set including biofilm thickness, equipment vibration acceleration, and ambient hydrogen sulfide concentration.

3. The method for predicting coal mine equipment failure based on big data according to claim 1, characterized in that: The S2 specifically includes: S21: Based on the db4 wavelet basis function, the electrochemical noise signal in the corrosion multimodal data set is decomposed at multiple levels. In the wavelet decomposition, each layer of the signal is split into a low-frequency part and a high-frequency part to form multiple sub-bands. The decomposition process continues for eight levels, so that the signal is divided into multiple frequency ranges. S22: Calculate the wavelet entropy value of each sub-band. After obtaining the target sub-band signal, analyze the energy distribution characteristics, use wavelet entropy analysis to quantify the complexity of the signal, determine whether the signal includes corrosion characteristics, count the proportion of different energy components in each sub-band, obtain the energy distribution characteristics, calculate the wavelet entropy value based on the energy distribution characteristics, measure the randomness and structural information of the signal, the larger the wavelet entropy value, the higher the signal complexity, set the entropy value threshold in all sub-band signals, when the wavelet entropy value of a sub-band exceeds the entropy value threshold, determine that the current sub-band includes effective microbial corrosion information, and retain it for subsequent feature extraction; S23: adaptively calculate the signal strength threshold of each sub-band to adapt it to signal changes in different environments. When the actual signal strength exceeds the signal strength threshold, it is determined to be a corrosion current pulse. Among all signal segments exceeding the signal strength threshold, key pulse parameters related to the corrosion process are extracted, including: Pulse amplitude, used to characterize the intensity of the corrosion reaction; Pulse duration, reflecting the length of time of the corrosion reaction; Pulse rise time, which is used to measure the suddenness of the corrosion process; S24: After extracting all key pulse parameters, a three-dimensional feature vector is constructed. The three-dimensional feature vector includes the average amplitude of the pulse, the total duration, and the change characteristics of the rise time. The corrosion activity is classified using the K-means clustering algorithm. The corrosion signal is divided into low activity level, medium activity level, and high activity level. The level division is based on the pulse amplitude and duration indicators. The clustering results are mapped to the time-frequency dimension to form a corrosion activity fingerprint spectrum. In the corrosion activity fingerprint spectrum, the horizontal axis represents time, the vertical axis represents characteristic frequency, and the color change reflects the different levels of corrosion activity.

4. The method for predicting coal mine equipment failure based on big data according to claim 3 is characterized in that: In the calculation of the wavelet entropy value of each sub-band in S22, for each sub-band , its wavelet entropy value Calculated as: ,in, For the Tier Energy proportion of each node: ,in, Indicates Tier The energy of a node is When , the layer is judged to contain effective corrosion features. is the entropy threshold.

5. The method for predicting coal mine equipment failure based on big data according to claim 4 is characterized in that: The signal strength threshold Calculated as: ,in, For the The standard deviation of the layer signal, is the number of sampling points of this layer.

6. The method for predicting coal mine equipment failure based on big data according to claim 1, characterized in that: The S3 specifically includes: S31: Construct a heterogeneous graph structure to describe the association between corrosion nodes and vibration nodes. The corrosion node is a pulse feature vector extracted from the corrosion activity fingerprint spectrum; the vibration node is a frequency band energy entropy vector extracted from the vibration time-frequency spectrum. After the nodes are constructed, define the association between corrosion and vibration, and use a weighting mechanism based on time difference to assign an edge weight to each corrosion-vibration connection edge. The value of the edge weight decays as the time difference between the corrosion node and the vibration node increases, ensuring that events that are close in time have higher correlation. S32: Construct a dual-channel graph attention network based on a heterogeneous graph structure, including a corrosion feature enhancement channel and a vibration-corrosion coupling channel, where: The corrosion feature enhancement channel is used to analyze the mutual correlation between corrosion nodes. The multi-head attention mechanism is used to calculate the attention of each corrosion node to other corrosion nodes, and the feature expression of the corrosion nodes is adjusted. Each corrosion node calculates its own correlation with other corrosion nodes, and uses the attention weighting method to update its own features to enhance the corrosion features. The vibration-corrosion coupling channel is used to analyze the impact of vibration signals on the corrosion process. The gating mechanism is used to screen effective associated edges. The gating mechanism calculates a dynamic weight based on the joint characteristics of corrosion and vibration to correct the initial time-related edge weight. The corrected edge weight is used to adjust the impact of vibration on the corrosion node. S33: Using singular value decomposition method to reduce the dimension of embedded features of corrosion-vibration nodes and decouple coupling effects, including calculating the mutual information between corrosion and vibration in the feature space after dimension reduction. Mutual information measures the statistical correlation between vibration signal and corrosion signal. The larger the value, the stronger the coupling between the two. When the mutual information exceeds the predetermined threshold, it indicates that the vibration signal and the corrosion signal have a strong dependence relationship, indicating that the corrosion process aggravates the fatigue vibration of the equipment. S43: After decoupling the relationship between corrosion and vibration, a corrosion-fatigue damage correlation map is constructed to describe the coupling mode between the two. The correlation map includes a horizontal axis and a vertical axis, where: The horizontal axis of the spectrum represents the corrosion activity intensity, which is derived from the chromaticity information of the corrosion activity fingerprint spectrum; The vertical axis of the spectrum represents the vibration energy entropy, which comes from the integration result of the time-frequency spectrum.

7. The method for predicting coal mine equipment failure based on big data according to claim 5, characterized in that: The edge weight is defined as: ,in, and Represent the timestamps of corrosion pulse events and vibration events, respectively. is the time-related scale factor; The corrosion feature enhancement channel uses a multi-head attention mechanism to calculate the spatial correlation between corrosion nodes: ,in, and is the learnable parameter matrix, is the feature dimension, is the attention weight between the eroded nodes, indicating The corrosion node is The attention scores of the eroded nodes are used to calculate the spatial correlation between nodes. Represents a corrosion node The characteristic vector of Represents a corrosion node The eigenvector of The vibration-corrosion coupling channel screens valid correlation edges through the gated attention mechanism: ； ； in, is the Sigmoid function, is the gating weight matrix; represents the vector concatenation operation, is the gated attention value of the vibration-corrosion coupling channel, is the corrected edge weight, represents the pulse feature vector, Represents the frequency band energy entropy vector.

8. The method for predicting coal mine equipment failure based on big data according to claim 1, characterized in that: The S4 specifically includes: S41: Construct a chemical corrosion rate prediction channel; extract the average amplitude of the pulse current from the corrosion activity fingerprint spectrum, obtain the hydrogen sulfide concentration gradient and the biofilm thickness growth rate, use the time convolution network to learn the dynamic characteristics of chemical corrosion, capture the corrosion change trend on different time scales, and establish a mapping relationship between the average amplitude, hydrogen sulfide concentration gradient and corrosion rate. Among them, the pulse current amplitude is used to measure the basic intensity of corrosion, the hydrogen sulfide concentration gradient affects the nonlinear change of the corrosion reaction rate, and the biofilm thickness growth rate is used to reflect the contribution of microbial activity to corrosion; S42: Construct a mechanical crack growth rate prediction channel; extract the vibration-corrosion mutual information from the corrosion-vibration correlation map, obtain the maximum principal stress and stress intensity factor amplitude to characterize the direct effect of mechanical load on crack growth. In the crack growth prediction, the traditional Paris law is used and the acceleration effect of corrosion on crack growth is considered; S43: Composite failure trajectory synthesis; establish the mapping relationship between the time scale of chemical corrosion and the number of mechanical cycles of mechanical fatigue, introduce time axis alignment processing, associate the evolution process of chemical corrosion with the expansion process of fatigue cracks, and analyze multiple failure modes in the same time coordinate system. After time axis alignment processing, calculate the composite failure trajectory curve, which describes the cumulative damage of corrosion and crack expansion during service. Use the calculation method of equivalent service time, combine chemical corrosion time and mechanical cycle number, and define the composite damage degree; S44: Early warning mechanism; after obtaining the composite damage degree, the early warning mechanism is set according to different damage thresholds. When the damage degree exceeds the set damage threshold but still does not reach the critical value, the first-level early warning is triggered and planned maintenance is recommended. If the damage degree continues to increase and its growth rate exceeds the critical value, the second-level early warning is triggered and emergency shutdown measures are taken.

9. The method for predicting coal mine equipment failure based on big data according to claim 8, characterized in that: In the early warning mechanism of S44: when When the alarm is triggered, the first level warning is triggered; When the following two conditions are met, the second level warning is triggered: ,in, represents the composite damage degree, represents the composite damage degree change rate, It's hours, is the equivalent service time.

10. A coal mine equipment fault prediction system based on big data, used to implement the coal mine equipment fault prediction method based on big data as described in any one of claims 1 to 9, characterized in that: Includes the following modules: Data acquisition module: The biofilm thickness on the equipment surface is collected through ultrasonic sensors, and the equipment vibration acceleration and environmental hydrogen sulfide concentration are simultaneously acquired to construct a corrosion multimodal data set; Signal processing module: used to perform electrochemical noise separation on corrosion multimodal data sets, and use wavelet entropy algorithm to extract microbial corrosion characteristic current pulses to generate corrosion activity fingerprint spectrum; Coupling analysis module: used to input the corrosion activity fingerprint spectrum into the graph attention network, decouple the coupling effect of mechanical vibration and microbial corrosion, and generate the corrosion-fatigue damage association map; Life prediction and early warning module: used to build a dual-channel life prediction model based on the corrosion-fatigue damage correlation map, calculate the composite failure trajectory curve of chemical corrosion rate and mechanical crack growth rate, and set a dynamic threshold early warning mechanism based on the composite failure trajectory curve to achieve coal mine equipment failure prediction and risk early warning.

Images