A coal mine equipment fault diagnosis method and related device

Abstract

The application discloses a coal mine equipment fault diagnosis method and related device, relates to the technical field of coal mine equipment fault diagnosis, and the method comprises the following steps: obtaining historical coal mine equipment data and preprocessing; inputting the preprocessed historical coal mine equipment data into a coal mine equipment fault diagnosis model, calculating a loss according to a coal mine equipment fault diagnosis result output by the coal mine equipment fault diagnosis model and a corresponding real label, optimizing model parameters based on the loss, and obtaining a trained coal mine equipment fault diagnosis model after iterative training; wherein the coal mine equipment fault diagnosis model is a neural network model based on a dynamic kernel principal component constraint mechanism and a quantum decoherence suppression mechanism; obtaining target real-time coal mine equipment data, and performing fault diagnosis on the target real-time coal mine equipment data by using the trained coal mine equipment fault diagnosis model to obtain a final coal mine equipment fault diagnosis result. The application can improve the accuracy of coal mine equipment fault diagnosis.

Description

Technical Field

[0001] The present application relates to the technical field of coal mine equipment fault diagnosis, and in particular to a coal mine equipment fault diagnosis method and related devices. Background Art

[0002] Coal mining equipment plays a crucial role in the production process, and its operating status directly impacts production safety and the lives of miners. Therefore, fault diagnosis of coal mining equipment has long been a crucial research topic within the industry. Traditional methods for fault diagnosis of coal mining equipment rely heavily on manual experience, rule-based judgment, or simple statistical analysis. These methods often fail to address the complex and volatile operating environment and diverse failure modes of coal mining equipment, resulting in low accuracy in coal mining equipment fault diagnosis. Summary of the Invention

[0003] The purpose of this application is to provide a coal mine equipment fault diagnosis method and related devices, which can effectively improve the accuracy of coal mine equipment fault diagnosis.

[0004] To achieve the above objectives, this application provides the following solutions.

[0005] In a first aspect, the present application provides a method for diagnosing coal mine equipment faults, which specifically includes the following steps.

[0006] Acquire historical coal mine equipment data; the historical coal mine equipment data refers to historical operation data of coal mine equipment, and each of the historical coal mine equipment data is marked with a real label reflecting the actual working status of the coal mine equipment.

[0007] The historical coal mine equipment data is preprocessed to obtain preprocessed historical coal mine equipment data.

[0008] The preprocessed historical coal mine equipment data is input into a coal mine equipment fault diagnosis model, and the loss is calculated based on the coal mine equipment fault diagnosis results output by the coal mine equipment fault diagnosis model and the corresponding true labels. The model parameters are optimized based on the loss, and a trained coal mine equipment fault diagnosis model is obtained after iterative training. The coal mine equipment fault diagnosis model is a neural network model based on a dynamic kernel principal component constraint mechanism and a quantum decoherence suppression mechanism. The dynamic kernel principal component constraint mechanism refers to dynamically adjusting the kernel function width according to the characteristics of different equipment fault types during training, and the quantum decoherence suppression mechanism refers to quantum noise suppression of weight updates and gradient calculations during training.

[0009] Target real-time coal mine equipment data is acquired, and fault diagnosis is performed on the target real-time coal mine equipment data using the trained coal mine equipment fault diagnosis model to obtain a final coal mine equipment fault diagnosis result; the target real-time coal mine equipment data refers to target operating data collected in real time and used for fault diagnosis of coal mine equipment.

[0010] Optionally, preprocessing the historical coal mine equipment data to obtain preprocessed historical coal mine equipment data specifically includes the following steps.

[0011] The historical coal mine equipment data is cleaned to obtain cleaned historical coal mine equipment data.

[0012] The historical coal mine equipment data after data cleaning is subjected to data normalization or standardization to obtain normalized or standardized historical coal mine equipment data.

[0013] The normalized or standardized historical coal mine equipment data is subjected to feature extraction in the time domain and the frequency domain to obtain features of the historical coal mine equipment data as preprocessed historical coal mine equipment data.

[0014] Optionally, the dynamic kernel principal component constraint mechanism specifically includes the following contents.

[0015] The Gaussian kernel function is used to perform coal mine equipment data mapping transformation, which is expressed as the following formula.

[0016] ;

[0017] in, is the kernel function, representing the input coal mine equipment data points and Projections in high-dimensional space; For the coal mine equipment data samples; For the Coal mine equipment data samples, characterization Neighboring coal mine equipment data samples.

[0018] The core width is dynamically adjusted according to the data characteristics of the coal mine equipment. The dynamic adjustment method of the core width is expressed as the following formula.

[0019] ;

[0020] in, For the The adjustment value of the kernel function width after iterations, which represents the adaptability of the kernel function; For the The adjustment value of the kernel function width after iterations, which represents the adaptability of the kernel function; is the square of the Euclidean distance between the input coal mine equipment data and the current mapping result, representing the difference between the coal mine equipment data; is the adjustment factor that controls the adjustment speed of the nuclear width; Control the flexibility of dynamic adjustment for preset tolerance error; The number of samples of coal mining equipment data input to the neural network for the current batch.

[0021] Optionally, the quantum decoherence suppression mechanism specifically includes the following contents.

[0022] When dynamically adjusting the kernel function width, the weight of the kernel function is updated using the quantum decoherence suppression mechanism. The decoherence effect of quantum noise is expressed as , and its updating process is expressed as follows.

[0023] ;

[0024] in, For the The kernel weight of the coal mine equipment data sample is The value at the iteration represents the weight after decoherence suppression; is the learning rate, which controls the amplitude of parameter updates; is the gradient of the objective function, which represents the direction of kernel function adjustment; is the quantum decoherence suppression term; It is a hyperparameter that controls the strength of noise suppression and characterizes the strength of quantum noise suppression.

[0025] The quantum decoherence suppression term is calculated using the following formula.

[0026] ;

[0027] in, is a symbolic function; is a parameter that controls the strength of inhibition; For the At the first iteration, the neural network neurons and The weight value of the neuron connection; is the decoherence attenuation factor.

[0028] The gradient calculation in back propagation is expressed as follows.

[0029] ;

[0030] in, is the gradient in back propagation; is the loss function of the neural network, which represents the error between the predicted value and the true value; is the L2 regularization term; is the weight of the neural network, which represents the connection strength between the input features of coal mine equipment data and neurons.

[0031] Under the quantum decoherence suppression mechanism, the gradient calculation is expressed as follows.

[0032] ;

[0033] in, is the gradient after quantum noise suppression, representing the weight update gradient adjusted by the quantum decoherence suppression mechanism; is the quantum noise model, which represents the noise suppression effect.

[0034] The updating process of coal mine equipment data features is expressed as the following formula.

[0035] ;

[0036] in, is the updated coal mine equipment data feature vector, representing the coal mine equipment data features after data feature mapping and quantum noise suppression; is the kernel function, representing the input coal mine equipment data points and Projections in high-dimensional space; is the coal mine equipment data feature vector after being mapped by the kernel function, representing the coal mine equipment data feature space after nonlinear transformation; for The corresponding quantum decoherence suppression term.

[0037] Optionally, target real-time coal mine equipment data is acquired, and fault diagnosis is performed on the target real-time coal mine equipment data using the trained coal mine equipment fault diagnosis model to obtain a final coal mine equipment fault diagnosis result, which specifically includes the following steps.

[0038] Obtain targeted real-time coal mine equipment data.

[0039] The target real-time coal mine equipment data is preprocessed to obtain preprocessed target real-time coal mine equipment data.

[0040] The pre-processed target real-time coal mining equipment data is input into the trained coal mining equipment fault diagnosis model to obtain a final coal mining equipment fault diagnosis result.

[0041] Optionally, preprocessing the target real-time coal mine equipment data to obtain the preprocessed target real-time coal mine equipment data specifically includes the following steps.

[0042] Data cleaning is performed on the target real-time coal mine equipment data to obtain cleaned target real-time coal mine equipment data.

[0043] The cleaned target real-time coal mine equipment data is subjected to data normalization processing or standardization processing to obtain normalized or standardized target real-time coal mine equipment data.

[0044] The normalized or standardized target real-time coal mine equipment data is subjected to time domain and frequency domain feature extraction to obtain target real-time coal mine equipment data features as preprocessed target real-time coal mine equipment data.

[0045] Optionally, after the step of obtaining the trained coal mine equipment fault diagnosis model, the coal mine equipment fault diagnosis method further comprises the following steps.

[0046] A model incremental training mechanism driven by perturbation awareness is used to continuously perform model incremental training on the trained coal mine equipment fault diagnosis model to obtain a model-incrementally-trained coal mine equipment fault diagnosis model; the model-incrementally-trained coal mine equipment fault diagnosis model is used as the trained coal mine equipment fault diagnosis model to perform fault diagnosis on the target real-time coal mine equipment data to obtain a final coal mine equipment fault diagnosis result.

[0047] Optionally, the model incremental training mechanism driven by perturbation awareness specifically comprises the following contents.

[0048] The distribution change of the input coal mine equipment data features is monitored in real time, and it is determined whether the distribution change of the coal mine equipment data features meets a feature distribution drift triggering condition to obtain a first determination result; the feature distribution drift triggering condition is as follows.

[0049] ;

[0050] wherein, represents the feature distribution drift triggering condition, is a Kullback-Leibler divergence; is a feature distribution of current coal mine equipment data; is a historical distribution during model training; is a preset triggering threshold.

[0051] When the first determination result is yes, it is determined that the distribution change of the coal mine equipment data features has feature distribution drift, and at this time, a local disturbance reconstruction mechanism is triggered to perform perturbation incremental updating on the affected hidden layer neurons, and each neuron weight that needs to be updated is represented as follows.

[0052] ;

[0053] in, is the original neural network connection weight; is the perturbation step length; is the perturbation gradient of the input feature, expressed as follows.

[0054] ;

[0055] in, It is the sliding mean of historical features, used to measure the feature deviation direction; is the regularization factor, used to control the disturbance scale; is a symbolic function; is the loss function of the neural network, which represents the error between the predicted value and the true value; For the Coal mine equipment data samples.

[0056] The memory-preserving regularization term is introduced to update the objective function. The updated objective function is expressed as follows.

[0057] ;

[0058] in, represents the updated objective function; is a snapshot of the key parameters of the original model; is the key parameter of the original neural network model; is the set of parameters that has the greatest impact on performance; is the retention strength factor.

[0059] In a second aspect, the present application proposes a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the coal mine equipment fault diagnosis method.

[0060] In a third aspect, the present application proposes a computer-readable storage medium having a computer program stored thereon, which implements the coal mine equipment fault diagnosis method when executed by a processor.

[0061] According to the specific embodiments provided in this application, this application has the following technical effects.

[0062] The present application provides a coal mine equipment fault diagnosis method and related device, which adopts a neural network model based on a dynamic kernel principal component constraint mechanism and a quantum decoherence suppression mechanism as the coal mine equipment fault diagnosis model. On the one hand, through the dynamic kernel principal component constraint mechanism, the kernel function width is dynamically adjusted according to the characteristics of different equipment fault types during the training process, which enables the coal mine equipment fault diagnosis model to adapt to different fault modes, effectively improves the generalization ability of the coal mine equipment fault diagnosis model, reduces interference between features, and improves the accuracy of coal mine equipment fault diagnosis. On the other hand, through the quantum decoherence suppression mechanism, quantum noise suppression is performed on weight updates and gradient calculations during the training process, which can effectively reduce the impact of quantum noise on model training and avoid noise amplification, thereby ensuring the stability and accuracy of the coal mine equipment fault diagnosis model and improving the accuracy of fault diagnosis of the coal mine equipment fault diagnosis model. Therefore, this application introduces a dynamic kernel principal component constraint mechanism and a quantum decoherence suppression mechanism during the model training process to continuously update the coal mine equipment data features during the training process, so that the coal mine equipment fault diagnosis model can better capture the diversity and complexity of coal mine equipment data, and ensure that the coal mine equipment fault diagnosis model can learn more accurate and effective features during the iterative training process, and obtain more accurate and reliable coal mine equipment fault diagnosis results. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0064] Figure 1 This is an application environment diagram of a coal mine equipment fault diagnosis method provided in one embodiment of the present application.

[0065] Figure 2 A flowchart of a method for diagnosing coal mine equipment faults provided in one embodiment of the present application.

[0066] Figure 3 This is a feature space distribution diagram of the traditional DNN method provided in one embodiment of the present application.

[0067] Figure 4 This is a feature space distribution diagram of the method of the present invention provided in one embodiment of the present application.

[0068] Figure 5 A schematic diagram of the noise reduction performance of the traditional DNN method provided in one embodiment of the present application.

[0069] Figure 6A schematic diagram of the noise reduction performance of the method of the present invention provided in one embodiment of the present application.

[0070] Figure 7 A schematic diagram of how the kernel width and gradient modulus vary with training rounds, provided in one embodiment of the present application.

[0071] Figure 8 A comparison diagram of the gradient noise distribution of the method of the present invention and the traditional DNN method provided in one embodiment of the present application.

[0072] Figure 9 A schematic diagram of the structure of a computer provided in one embodiment of the present application. DETAILED DESCRIPTION

[0073] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0074] With the advancement of sensor technology, coal mining equipment can now collect a large amount of operational data in real time through a variety of sensors (such as vibration, temperature, pressure, current, and speed). This data provides a crucial basis for fault diagnosis. Furthermore, the rapid development of machine learning has also provided new insights and methods for coal mining equipment fault diagnosis. Machine learning-based fault diagnosis methods can be trained using large amounts of historical data to automatically identify equipment failure modes and rapidly diagnose equipment anomalies, significantly improving diagnostic accuracy and efficiency. Traditional fault diagnosis methods lack the ability to effectively fusion features and process multi-dimensional data when dealing with multi-sensor data from coal mining equipment. Existing methods often rely on single-sensor data or inadequately fuse multi-sensor data, making it difficult to fully capture the equipment's operating status and potential failure modes. Therefore, effectively fusing the time and frequency domain features of multiple sensors to improve the accuracy of coal mining equipment fault diagnosis has become a pressing technical challenge in this field.

[0075] Furthermore, the types of failures in coal mine equipment are complex and diverse. Traditional machine learning methods often lack adaptability to different types of failures when processing equipment data, resulting in weak generalization capabilities of fault diagnosis models. Existing diagnostic systems often rely on static feature selection and fixed algorithm structures, failing to fully account for the diversity and complexity of coal mine equipment failure modes, resulting in low diagnostic accuracy for different equipment and fault types. Therefore, how to improve the model's adaptability to different fault types through a dynamic adjustment mechanism has become one of the key technical issues to be addressed in this application.

[0076] Traditional machine learning methods often fail to consider the impact of quantum noise on models during training. This is especially true when processing high-dimensional data and complex nonlinear features, where noise significantly impacts the stability and accuracy of model training. This noise interference can lead to unstable weight updates during model training, affecting the accuracy and reliability of fault diagnosis results. Therefore, this application addresses the question of how to effectively suppress the interference of quantum noise on the neural network training process, ensuring that training stability and accuracy are maintained even with high-dimensional data.

[0077] Traditional coal mining equipment fault diagnosis methods often suffer from overfitting during training. This is particularly true when faced with complex data and diverse fault modes. This often results in the model being unable to accurately identify the equipment's actual fault state, thus impacting the effectiveness of fault diagnosis. Therefore, designing a highly adaptive feature update and learning mechanism to prevent overfitting and improve the model's robustness and generalization capabilities is another key technical challenge addressed in this application.

[0078] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0079] The coal mine equipment fault diagnosis method provided in the embodiment of the present application can be applied to Figure 1The application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data required by the server 104 to process. The data storage system can be set up separately, or integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send historical coal mine equipment data and target real-time coal mine equipment data to the server 104. After the server 104 receives the historical coal mine equipment data and the target real-time coal mine equipment data, for the historical coal mine equipment data and the target real-time coal mine equipment data, the server 104 first preprocesses the historical coal mine equipment data; input the preprocessed historical coal mine equipment data into the coal mine equipment fault diagnosis model, calculate the loss according to the coal mine equipment fault diagnosis result output by the coal mine equipment fault diagnosis model and the corresponding true label, optimize the model parameters based on the loss, and get the trained coal mine equipment fault diagnosis model after iterative training; wherein, the coal mine equipment fault diagnosis model is a neural network model based on dynamic kernel principal component constraint mechanism and quantum decoherence suppression mechanism; then, for the target real-time coal mine equipment data, the trained coal mine equipment fault diagnosis model is used to diagnose the target real-time coal mine equipment data, and the final coal mine equipment fault diagnosis result is obtained. The server 104 can feed back the final coal mine equipment fault diagnosis result to the terminal 102. In addition, in some embodiments, the coal mine equipment fault diagnosis method can also be implemented by the server 104 or the terminal 102 alone, such as the terminal 102 can directly diagnose the historical coal mine equipment data and the target real-time coal mine equipment data, or the server 104 can obtain the historical coal mine equipment data and the target real-time coal mine equipment data from the data storage system, and diagnose the historical coal mine equipment data and the target real-time coal mine equipment data.

[0080] Among them, the terminal 102 can be but not limited to various desktop computers, notebook computers, smart phones, tablet computers and Internet of Things devices, and the server 104 can be implemented by an independent server or a server cluster composed of multiple servers, and can also be a cloud server.

[0081] In an exemplary embodiment, as shown in Figure 2 A coal mine equipment fault diagnosis method is provided, which is executed by a computer device, specifically by a terminal or a server, or by a terminal and a server together. In the embodiment of the present application, the method is applied to the server 104 in Figure 1 The following steps S1 to S4 are included.

[0082] S1: Obtain historical coal mine equipment data.

[0083] The historical coal mine equipment data refers to the historical operating data of coal mine equipment. This data is used in the model training phase to train the coal mine equipment fault diagnosis model and obtain a trained coal mine equipment fault diagnosis model. Each piece of historical coal mine equipment data is annotated with a real-world label that reflects the actual operating status of the coal mine equipment.

[0084] S2: Preprocessing the historical coal mine equipment data to obtain preprocessed historical coal mine equipment data.

[0085] In this embodiment, step S2 pre-processes the historical coal mine equipment data to obtain the pre-processed historical coal mine equipment data, which specifically includes the following steps.

[0086] S21: performing data cleaning on the historical coal mine equipment data to obtain the cleaned historical coal mine equipment data.

[0087] S22: performing data normalization or standardization on the historical coal mine equipment data after data cleaning to obtain normalized or standardized historical coal mine equipment data.

[0088] S23: performing time domain and frequency domain feature extraction on the normalized or standardized historical coal mine equipment data to obtain historical coal mine equipment data features as preprocessed historical coal mine equipment data.

[0089] S3: Input the preprocessed historical coal mine equipment data into a coal mine equipment fault diagnosis model, calculate the loss according to the coal mine equipment fault diagnosis results output by the coal mine equipment fault diagnosis model and the corresponding true labels, optimize the model parameters based on the loss, and obtain a trained coal mine equipment fault diagnosis model after iterative training is completed.

[0090] In this embodiment, the coal mine equipment fault diagnosis model is a neural network model based on a dynamic kernel principal component constraint mechanism and a quantum decoherence suppression mechanism. The dynamic kernel principal component constraint mechanism refers to dynamically adjusting the kernel function width according to the characteristics of different equipment fault types during training, and the quantum decoherence suppression mechanism refers to quantum noise suppression of weight updates and gradient calculations during training.

[0091] S4: Acquire target real-time coal mine equipment data, and use the trained coal mine equipment fault diagnosis model to perform fault diagnosis on the target real-time coal mine equipment data to obtain a final coal mine equipment fault diagnosis result.

[0092] The target real-time coal mine equipment data refers to the target operating data collected in real time and used for coal mine equipment fault diagnosis. This data is used in the actual application phase of the model, primarily to leverage the trained coal mine equipment fault diagnosis model to complete real-time coal mine equipment fault diagnosis. The final coal mine equipment fault diagnosis results are based on the target real-time coal mine equipment data.

[0093] In this embodiment, step S4 acquires target real-time coal mine equipment data, and uses the trained coal mine equipment fault diagnosis model to perform fault diagnosis on the target real-time coal mine equipment data to obtain the final coal mine equipment fault diagnosis result, which specifically includes the following steps.

[0094] S41: Acquire target real-time coal mine equipment data.

[0095] S42: Preprocessing the target real-time coal mine equipment data to obtain preprocessed target real-time coal mine equipment data.

[0096] S43: Inputting the pre-processed target real-time coal mining equipment data into the trained coal mining equipment fault diagnosis model to obtain a final coal mining equipment fault diagnosis result.

[0097] In this embodiment, step S42 pre-processes the target real-time coal mine equipment data to obtain the pre-processed target real-time coal mine equipment data, which specifically includes the following steps.

[0098] S421: performing data cleaning on the target real-time coal mine equipment data to obtain cleaned target real-time coal mine equipment data.

[0099] S422: performing data normalization or standardization on the target real-time coal mine equipment data after data cleaning to obtain normalized or standardized target real-time coal mine equipment data.

[0100] S423: Performing time domain and frequency domain feature extraction on the normalized or standardized target real-time coal mine equipment data to obtain target real-time coal mine equipment data features as pre-processed target real-time coal mine equipment data.

[0101] In an exemplary embodiment, after the step of obtaining a trained coal mine equipment fault diagnosis model after iterative training is completed in step S3, the coal mine equipment fault diagnosis method further includes the following steps.

[0102] The trained coal mine equipment fault diagnosis model is continuously incrementally trained using a perturbation-sensing driven model incremental training mechanism to obtain a coal mine equipment fault diagnosis model after incremental training. The incrementally trained coal mine equipment fault diagnosis model is used as the trained coal mine equipment fault diagnosis model to perform fault diagnosis on the target real-time coal mine equipment data to obtain a final coal mine equipment fault diagnosis result.

[0103] In this embodiment, the perturbation perception-driven model incremental training mechanism specifically includes the following contents.

[0104] The distribution changes of the input coal mine equipment data characteristics are monitored in real time to determine whether the distribution changes of the coal mine equipment data characteristics meet the characteristic distribution drift triggering condition, and obtain a first judgment result, including the following two situations.

[0105] (1) When the first judgment result is yes, it is determined that the distribution change of the coal mine equipment data feature triggers the feature distribution drift, and the local perturbation reconstruction mechanism is triggered at this time, and the affected hidden layer neurons are incrementally updated with perturbations, and the memory retention regularization term is introduced to update the objective function. The updated objective function is used to train the coal mine equipment fault diagnosis model after the model incremental training, and the coal mine equipment fault diagnosis model after the model incremental training is used as the trained coal mine equipment fault diagnosis model to perform coal mine equipment fault diagnosis.

[0106] (2) When the first judgment result is no, it is determined that the distribution change of the coal mine equipment data characteristics has not caused feature distribution drift. At this time, the perturbation perception-driven model incremental training mechanism is not triggered, and the trained coal mine equipment fault diagnosis model is still used to perform coal mine equipment fault diagnosis.

[0107] In order to make the technical solution of this embodiment clearer, the specific implementation process of the technical solution of this embodiment is described in detail below in the form of examples, including the following implementation steps.

[0108] S1: Collection and annotation of historical coal mine equipment data.

[0109] Coal mine equipment is monitored in real time, and data from coal mine equipment over a period of time is collected as historical coal mine equipment data for model training, including various sensor data such as vibration, temperature, pressure, current, and speed. Each sensor will record the operating data of the coal mine equipment in real time, and these historical coal mine equipment data need to be annotated with real labels based on the actual working status of the coal mine equipment. The annotated content includes normal operating status, different types of fault status, and corresponding fault levels.

[0110] S2: Preprocessing of historical coal mine equipment data.

[0111] First, historical coal mine equipment data is cleaned to remove noise and outliers generated during the acquisition process and ensure accuracy. Next, data normalization or standardization is required to enable comparison and analysis of sensor data from different dimensions on the same scale. For multi-dimensional time series data, feature extraction techniques are used to extract meaningful features from the time and frequency domains, such as mean, variance, peak value, kurtosis, and spectral energy. This converts the raw data into feature vectors that can be used for model training.

[0112] S3: Model training of coal mine equipment fault diagnosis model.

[0113] This embodiment uses a neural network model as a coal mine equipment fault diagnosis model, and its training process is as follows.

[0114] S301: Initialize the parameters of the neural network.

[0115] At the beginning of training, the structure of the neural network is determined, including the number of input layers, hidden layers, and output layers, as well as the number of neurons in each layer.

[0116] In an exemplary embodiment, the structure of the neural network is as follows: the input layer contains coal mine equipment data features (such as vibration, temperature, pressure, etc.) of multiple sensor data, the hidden layer contains multiple neurons for extracting the characteristic relationship of coal mine equipment data, and the output layer corresponds to the fault category of the coal mine equipment. The specific number of layers and neurons will be adjusted according to the complexity of the coal mine equipment and the specific requirements of the task. For example, the input layer has 10 neurons, corresponding to the coal mine equipment data features collected by 10 sensors; there are 2 hidden layers, the first hidden layer has 20 neurons, and the second hidden layer has 15 neurons; the output layer has 3 neurons, corresponding to 3 types of coal mine equipment faults (such as mechanical faults, electrical faults, and hydraulic faults).

[0117] The number of hidden layers can be multiple, for example, 5 layers.

[0118] During the training process of this embodiment, the network weights and bias terms are randomly initialized to avoid symmetry destruction in the initial stage, and then the gradient descent algorithm is used for optimization.

[0119] In view of the characteristics of coal mine equipment data, the input coal mine equipment data is multi-dimensional coal mine equipment data collected by multiple sensors. Each coal mine equipment data sample contains different sensor data features (such as vibration, temperature, pressure, etc.). In one embodiment, the coal mine equipment data is time series data collected by six types of sensors (vibration triaxial accelerometer, temperature sensor, pressure sensor, current sensor, speed sensor, oil particle sensor). Each sensor extracts five time domain data features (mean, variance, peak, kurtosis, and waveform factor) and three frequency domain data features (main frequency amplitude, spectrum energy, and frequency band entropy) within a 1-second time window to form a 6×8=48-dimensional coal mine equipment data feature vector.

[0120] In this embodiment, the connection weights and bias terms in the neural network are randomly initialized, and an adaptive dynamic kernel principal component constraint mechanism is set according to the distribution of the training coal mine equipment data. The initialization of the neural network includes randomly initializing the weights and bias terms, which is expressed as the following formula.

[0121] (1).

[0122] (2).

[0123] in, is the weight of the neural network, representing the connection strength between the input features of coal mine equipment data and neurons; The bias of the neural network represents the offset of the neuron activation function; Indicates that the weight initialization follows the mean of 0 and the variance of to ensure the symmetry of the model.

[0124] In this embodiment, during the initialization process, the parameters of the kernel function need to be initialized, and a dynamic adjustment mechanism corresponding to the dynamic kernel principal component constraint mechanism is designed. In one embodiment, a Gaussian kernel function is used to perform coal mine equipment data mapping transformation, which is expressed as the following formula.

[0125] (3).

[0126] in, is the kernel function, representing the input coal mine equipment data points and Projections in high-dimensional space; For the coal mine equipment data samples; For the Coal mine equipment data samples, characterization Neighboring coal mine equipment data samples.

[0127] At this time, the kernel width is dynamically adjusted according to the coal mine equipment data characteristics of the coal mine equipment to adapt to different equipment failure types. During the training process, the dynamic adjustment method of the kernel width is expressed as the following formula.

[0128] (4).

[0129] in, For the The adjustment value of the kernel function width after iterations, which represents the adaptability of the kernel function; For the coal mine equipment data samples; For the Coal mine equipment data samples, characterization Neighboring coal mine equipment data samples; For the The adjustment value of the kernel function width after iterations, which represents the adaptability of the kernel function; is the square of the Euclidean distance between the input coal mine equipment data and the current mapping result, representing the difference between the coal mine equipment data; is the adjustment factor that controls the adjustment speed of the nuclear width; Control the flexibility of dynamic adjustment for preset tolerance error; The number of samples of coal mining equipment data input to the neural network in the current batch. Set to 0.2, Set to 0.00001.

[0130] S302: Coal mine equipment data feature constraints and adaptive learning.

[0131] During training, each input feature of coal mining equipment data is nonlinearly mapped using a dynamic kernel principal component constraint mechanism, ensuring that the coal mining equipment data feature space is more adaptable to the diversity and complexity of coal mining equipment failures. Each input vector is mapped from the coal mining equipment data feature space to a higher-dimensional space using a kernel function, allowing the neural network to learn more complex relationships between coal mining equipment data features.

[0132] At the same time, the coal mine equipment data feature mapping is represented by the kernel function to reduce the interference caused by quantum noise and ensure the stability of the coal mine equipment data feature transformation process, which is expressed as the following formula.

[0133] (5).

[0134] in, is the coal mine equipment data feature vector after mapping by kernel function (corresponding to coal mine equipment data samples), representing the feature space of coal mine equipment data after nonlinear transformation; is the input of the function; For the The weight of the kernel function of each coal mine equipment data sample represents the contribution of each training coal mine equipment data sample to the coal mine equipment data feature space; is the kernel function, representing the input coal mine equipment data points and Projection in high-dimensional space.

[0135] In this embodiment, during dynamic adjustment, the weight of the kernel function is updated through the quantum decoherence suppression mechanism. Assume that the decoherence effect of quantum noise can be expressed as , and its updating process is expressed as follows.

[0136] (6).

[0137] in, For the The kernel weight of the coal mine equipment data sample is The value at the iteration represents the weight after decoherence suppression; is the learning rate, which controls the amplitude of parameter updates; is the gradient of the objective function, which represents the direction of kernel function adjustment; is the quantum decoherence suppression term, which reduces the impact of noise; It is a hyperparameter that controls the strength of noise suppression and characterizes the strength of quantum noise suppression.

[0138] In this embodiment, the design of the quantum decoherence suppression term makes small weights more strongly suppressed when the weights are updated, preventing noise amplification while retaining the learning ability of key coal mine equipment data characteristics. The calculation method is expressed as the following formula.

[0139] (7).

[0140] in, is the decoherence attenuation factor (default is 0.1), which suppresses quantum noise interference with large weight values ​​through exponential decay; is a sign function that preserves the gradient direction, is a parameter that controls the strength of inhibition; For the At the first iteration, the neural network neurons and The weights of the neuron connections.

[0141] S303: Back propagation and gradient adjustment.

[0142] During the backpropagation process, the traditional gradient calculation method will be affected by the quantum decoherence suppression mechanism. To ensure the accuracy and stability of the gradient calculation, the weight update of the neural network is optimized, and the convergence speed and accuracy of the training process are improved. The gradient calculation method in backpropagation is expressed as follows.

[0143] (8).

[0144] in, is the gradient in back propagation, is the loss function of the neural network, which represents the error between the predicted value and the true value; It is the L2 regularization term to prevent overfitting; is the weight of the neural network, which represents the connection strength between the input features of coal mine equipment data and neurons.

[0145] In this embodiment, under the quantum decoherence suppression mechanism, the calculation method of the gradient is adjusted to the following formula.

[0146] (9).

[0147] in, is the gradient after quantum noise suppression, representing the weight update gradient adjusted by the quantum decoherence suppression mechanism; is the quantum noise model, which represents the noise suppression effect; It is a hyperparameter for controlling the noise suppression strength, and characterizes the strength of quantum noise suppression. Set to 0.01.

[0148] The suppression of quantum noise is not only reflected in the adjustment of the kernel function, but also in the update process of each weight and gradient.

[0149] This embodiment introduces a quantum decoherence suppression mechanism in the back propagation of the neural network, which can reduce the impact of noise on the network training process, ensure more accurate model updates, and ultimately improve the accuracy of fault diagnosis of the coal mine equipment fault diagnosis model.

[0150] S304: Update and reorganize coal mine equipment data features.

[0151] During the training process of the neural network, as the coal mining equipment data features of each layer are continuously updated and reorganized, the network gradually learns a more effective representation of the coal mining equipment data features.

[0152] At this time, through the dynamic kernel principal component constraint mechanism, the coal mine equipment data characteristics will be continuously mapped to a higher-dimensional coal mine equipment data feature space in order to better capture the nonlinear relationship in the coal mine equipment data.

[0153] At the same time, the quantum decoherence suppression mechanism plays a role in this process, reducing the impact of noise on the update of coal mining equipment data features and ensuring that each neuron can effectively learn the core features of coal mining equipment data. Therefore, the dynamic kernel principal component constraint mechanism and the quantum decoherence suppression mechanism are synchronized and mutually coordinated. Through the synergistic cooperation between the dynamic kernel principal component constraint mechanism and the quantum decoherence suppression mechanism, the precision of the coal mining equipment fault diagnosis model can be effectively improved, and the accuracy of coal mining equipment fault diagnosis can be enhanced.

[0154] Based on this, the updating process of coal mine equipment data features is expressed as the following formula.

[0155] (10).

[0156] in, is the updated coal mine equipment data feature vector, representing the coal mine equipment data features after data feature mapping and quantum noise suppression; is the coal mine equipment data feature vector after being mapped by the kernel function, representing the coal mine equipment data feature space after nonlinear transformation; for The corresponding quantum decoherence suppression term represents the reduction of the influence of characteristic noise in coal mine equipment data through quantum noise suppression.

[0157] S305: Multi-level integration and final output stage.

[0158] After multiple rounds of optimization and coal mine equipment data feature processing, the neuron information of all layers is finally integrated into the output layer of the neural network.

[0159] The output layer of the neural network uses a specific activation function to perform nonlinear transformation and generate the final classification results. During this process, the dynamic kernel principal component constraint mechanism and the quantum decoherence suppression mechanism work together to ensure that the neural network algorithm accurately models complex relationships. The calculation formula for the output result is as follows.

[0160] (11).

[0161] in, It is the final prediction output, representing the prediction results of the coal mine equipment fault diagnosis model; is the Sigmoid activation function, which represents the nonlinear transformation process; is the weight of the neural network, which represents the connection strength between input and output; is the updated coal mine equipment data feature vector, representing the coal mine equipment data features after data feature mapping and quantum noise suppression; It is the bias term of the neural network, representing the activation offset of the neuron.

[0162] S306: Repeat the above steps in the iteration step S3 until the preset stop iteration condition is met, which means that the model training is completed. In one embodiment, the preset stop iteration condition is reaching a preset maximum number of iterations, and the preset maximum number of iterations is set to 1000 times.

[0163] S4: Coal mine equipment fault diagnosis.

[0164] After the coal mine equipment fault diagnosis model training is completed, the target real-time coal mine equipment data is obtained according to the method of step S1, and the target real-time coal mine equipment data is preprocessed according to the method of step S2. The trained coal mine equipment fault diagnosis model is used to perform coal mine equipment fault diagnosis on the preprocessed target real-time coal mine equipment data to obtain the final coal mine equipment fault diagnosis result.

[0165] S5: Model incremental training.

[0166] In response to new fault types or operating condition changes that occur in the coal mine equipment operating environment, this embodiment proposes a perturbation-sensing driven model incremental training mechanism to achieve continuous learning and online adaptation of the coal mine equipment fault diagnosis model in practical applications.

[0167] In some scenarios, it is not necessary to fully train the original coal mine equipment fault diagnosis model. After the coal mine equipment fault diagnosis model is deployed, the system monitors the distribution changes of the input coal mine equipment data features in real time. If the distribution changes of the input coal mine equipment data features are detected to have feature distribution drift, the triggering condition of the feature distribution drift is as follows.

[0168] (12).

[0169] in, Indicates the triggering condition for feature distribution drift; is the Kullback-Leibler divergence, which measures the difference between the distribution of new and old input data; is the characteristic distribution of current coal mine equipment data; is the historical distribution during model training; is the preset trigger threshold.

[0170] When it is determined that the distribution change of coal mine equipment data features has caused feature distribution drift, the system will automatically trigger the perturbation perception-driven model incremental training mechanism. The perturbation perception-driven model incremental training mechanism is a local perturbation reconstruction mechanism that is used to perform perturbation incremental updates on the affected hidden layer neurons after sensing feature distribution drift. The weight of each neuron that needs to be updated is expressed as the following formula.

[0171] (13).

[0172] in, is the original neural network connection weight, is the perturbation step length; is the perturbation gradient of the input feature, defined as follows.

[0173] (14).

[0174] in, It is the sliding mean of historical features, used to measure the feature deviation direction; is a regularization factor used to control the disturbance scale.

[0175] In this embodiment, in order to prevent the model from catastrophic forgetting under new tasks, a memory preservation regularization term is introduced in the incremental training stage, and its objective function is updated as follows.

[0176] (15).

[0177] in, represents the updated objective function; A snapshot of the key parameters of the original neural network model; is the key parameter of the original neural network model; The set of parameters that have the greatest impact on performance is screened using the Fisher information matrix; is the retention strength factor.

[0178] This embodiment inputs real-time monitored target coal mine equipment data into a trained coal mine equipment fault diagnosis model. The trained coal mine equipment fault diagnosis model then predicts and classifies the equipment's operating status based on the learned characteristics and patterns of the coal mine equipment data. Specifically, the trained coal mine equipment fault diagnosis model performs a nonlinear transformation on the input coal mine equipment data characteristics. After processing through a multi-layer neural network, it ultimately outputs the equipment's fault category. Using the activation function of the output layer, the trained coal mine equipment fault diagnosis model accurately determines whether the coal mine equipment is currently in a fault state.

[0179] Traditional neural network models typically rely on fixed feature extraction and linear mapping when diagnosing coal mining equipment faults. This embodiment, by employing a dynamic kernel principal component constraint mechanism, achieves nonlinear mapping and dynamically adjusts the kernel function width based on the characteristics of different equipment fault types, adapting to different fault modes. This mechanism effectively improves the model's generalization capability and reduces interference between features. Furthermore, during model training, a quantum decoherence suppression mechanism effectively reduces the impact of quantum noise on model training. By suppressing quantum noise during weight updates and gradient calculations during training, noise amplification is avoided, ensuring the stability and precision of the coal mining equipment fault diagnosis model and improving the accuracy of coal mining equipment fault diagnosis. The feature mapping of this embodiment not only relies on static feature extraction but also continuously updates coal mining equipment data features during training through a dynamic kernel principal component constraint mechanism and quantum decoherence suppression mechanism. This adaptive learning mechanism better captures the diversity and complexity of coal mining equipment data, ensuring that the coal mining equipment fault diagnosis model learns more accurate and effective features over multiple rounds of iterations. This embodiment combines multiple sensor data (such as vibration, temperature, pressure, and current) with time-domain and frequency-domain features. By integrating multi-dimensional feature data, this ensures that the coal mining equipment fault diagnosis model can fully capture the operating status and potential faults of coal mining equipment, enabling more accurate fault diagnosis. Furthermore, this embodiment optimizes the gradient calculation during the backpropagation process by employing a quantum noise suppression mechanism, thereby ensuring the accuracy and convergence speed of the coal mining equipment fault diagnosis model during training. This effectively reduces noise interference during training, ensuring that each weight update is carried out in the correct direction, and thus ensuring the accuracy of coal mining equipment fault diagnosis.

[0180] In order to verify the effectiveness of the method of the present invention, the following comparative experiments were conducted between the method of the present invention and the traditional DNN (Deep Nueral Network) method.

[0181] 1) By visualizing the distribution of three-dimensional feature space, we compared the feature extraction capabilities of the proposed method with those of traditional DNN methods. To verify the optimization effect of the dynamic kernel principal component constraint mechanism on the feature space structure, we focused on the separability of different fault types in the feature space. Figure 3 This is the feature space distribution diagram of the traditional DNN method. Figure 4The figure below is the feature space distribution diagram of the method of the present invention. Data points of different colors represent data points that are not in the same plane in the three-dimensional space. The darker the color of the data point, the closer the feature space distribution position of the data point is to the bottom layer. It can be seen that the features extracted by the method of the present invention present a clear temporal spiral structure. Data points of different fault categories form a clearly separated cluster distribution in the three-dimensional space, while the feature distribution of the traditional method presents a disordered random scattered state. This difference intuitively demonstrates that the dynamic kernel principal component constraint mechanism can transform the original sensor data into a highly separable feature representation through nonlinear mapping, laying a solid foundation for subsequent classification tasks.

[0182] 2) A three-dimensional response surface was used to compare the performance stability of different methods in a complex noise environment. The experiment focused on verifying the dual resistance of the quantum decoherence suppression mechanism to sensor noise and quantum noise, simulating the complex noise interference in industrial scenarios. Figure 5 Schematic diagram of the noise resistance performance of traditional DNN methods. Figure 6 This figure illustrates the noise-reduction performance of the proposed method. The results show that as the intensities of both types of noise increase simultaneously, the classification accuracy of the proposed method decreases only gently, with its response surface exhibiting a stable plateau. In contrast, the performance of the conventional method degrades dramatically as the noise increases, forming a steep downward gradient. This demonstrates that the quantum decoherence suppression mechanism can effectively separate noise signals from true features, maintaining reliable feature learning capabilities even in strong interference environments.

[0183] 3) The dynamic parameter change curve reveals the co-evolutionary law of the kernel function adjustment mechanism and the training process. The experiment tracks the real-time changes of the kernel width parameter and the gradient modulus during the training process, aiming to clarify the role of the dynamic adjustment mechanism in improving training stability. Figure 7 As shown in the figure, the kernel width parameter exhibits regular oscillation attenuation in the early and middle stages of training, which forms a precise match with the gradient convergence process and finally stabilizes in the theoretical expected range. This adaptive adjustment characteristic shows that the kernel function parameters can automatically optimize the mapping scale according to the feature learning progress, avoiding both unstable fluctuations in the early training and excessive contraction in the later training, and achieving dynamic matching between the feature space scale and the model convergence state.

[0184] 4) The anti-interference capabilities of different methods are quantitatively compared through the gradient noise distribution histogram. The experiment focuses on the quantum noise interference during neural network training and analyzes the statistical distribution characteristics of the gradient update amount. Figure 8This is a comparison diagram of the gradient noise distribution of the method of the present invention and the traditional DNN method. It can be seen that the gradient noise generated by the method of the present invention is concentrated in a narrow range near the zero value, presenting a symmetrical sharp peak shape; the noise distribution of the traditional method shows a bimodal diffusion feature. This distribution difference directly reflects the effect of the quantum decoherence suppression mechanism. The dynamic weight decay strategy is used to suppress abnormal gradient pulses and filter high-frequency noise components, thereby ensuring that parameter updates are steadily advanced in the correct direction, significantly improving the robustness of the training process.

[0185] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 9 As shown. The computer device includes a processor, a memory, an input / output interface (I / O), and a communication interface. The processor, memory, and I / O interface are connected via a system bus, and the communication interface is connected to the system bus via the I / O 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 computer program in the non-volatile storage medium. The database of the computer device is used to store historical coal mine equipment data and target real-time coal mine equipment data. The I / O interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method for diagnosing coal mine equipment faults is implemented.

[0186] Those skilled in the art will understand that Figure 9 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0187] In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0188] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0189] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. 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 above-mentioned embodiments. In particular, any reference to memory, database, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can 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 can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0190] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0191] This document uses specific examples to illustrate the principles and implementation methods of this application. The description of the above examples is only intended to help understand the method and core concept of this application. At the same time, for those skilled in the art, based on the concept of this application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for diagnosing coal mine equipment faults, characterized in that: The coal mine equipment fault diagnosis method comprises: Acquire historical coal mine equipment data; the historical coal mine equipment data refers to historical operation data of the coal mine equipment, each of the historical coal mine equipment data is annotated with a real label reflecting the actual working status of the coal mine equipment; Preprocessing the historical coal mine equipment data to obtain preprocessed historical coal mine equipment data; The preprocessed historical coal mine equipment data is input into a coal mine equipment fault diagnosis model, and the loss is calculated based on the coal mine equipment fault diagnosis results output by the coal mine equipment fault diagnosis model and the corresponding true labels. The model parameters are optimized based on the loss, and a trained coal mine equipment fault diagnosis model is obtained after iterative training. The coal mine equipment fault diagnosis model is a neural network model based on a dynamic kernel principal component constraint mechanism and a quantum decoherence suppression mechanism. The dynamic kernel principal component constraint mechanism refers to dynamically adjusting the kernel function width according to the characteristics of different equipment fault types during training, and the quantum decoherence suppression mechanism refers to quantum noise suppression of weight updates and gradient calculations during training. The dynamic kernel principal component constraint mechanism specifically includes: The Gaussian kernel function is used to map the coal mine equipment data, which can be expressed as: ； in, is the kernel function, representing the input coal mine equipment data points and Projections in high-dimensional space; For the coal mine equipment data samples; For the coal mine equipment data samples, characterizing Neighboring coal mine equipment data samples; The core width is dynamically adjusted according to the data characteristics of the coal mining equipment. The dynamic adjustment method of the core width is expressed as follows: ； in, For the The adjustment value of the kernel function width after iterations, which represents the adaptability of the kernel function; For the The adjustment value of the kernel function width after iterations, which represents the adaptability of the kernel function; is the square of the Euclidean distance between the input coal mine equipment data and the current mapping result, representing the difference between the coal mine equipment data; is the adjustment factor that controls the adjustment speed of the nuclear width; Control the flexibility of dynamic adjustment for preset tolerance error; The number of samples of coal mining equipment data input to the neural network for the current batch; Target real-time coal mine equipment data is acquired, and fault diagnosis is performed on the target real-time coal mine equipment data using the trained coal mine equipment fault diagnosis model to obtain a final coal mine equipment fault diagnosis result; the target real-time coal mine equipment data refers to target operating data collected in real time and used for fault diagnosis of coal mine equipment.

2. The coal mine equipment fault diagnosis method according to claim 1, characterized in that: Preprocessing the historical coal mine equipment data to obtain preprocessed historical coal mine equipment data specifically includes: performing data cleaning on the historical coal mine equipment data to obtain the cleaned historical coal mine equipment data; Performing data normalization or standardization on the historical coal mine equipment data after data cleaning to obtain normalized or standardized historical coal mine equipment data; The normalized or standardized historical coal mine equipment data is subjected to feature extraction in the time domain and the frequency domain to obtain features of the historical coal mine equipment data as preprocessed historical coal mine equipment data.

3. The coal mine equipment fault diagnosis method according to claim 1, characterized in that: The quantum decoherence suppression mechanism specifically includes: When dynamically adjusting the kernel function width, the weight of the kernel function is updated using the quantum decoherence suppression mechanism. The decoherence effect of quantum noise is expressed as , and its update process is expressed as: ； in, For the The kernel weight of the coal mine equipment data sample is The value at the iteration represents the weight after decoherence suppression; is the learning rate, which controls the amplitude of parameter updates; is the gradient of the objective function, which represents the direction of kernel function adjustment; is the quantum decoherence suppression term; To control the hyperparameter of noise suppression strength, characterize the strength of quantum noise suppression; The quantum decoherence suppression term is calculated using the following formula: ； in, is a symbolic function; is a parameter that controls the strength of inhibition; For the At the first iteration, the neural network neurons and The weight value of the neuron connection; is the decoherence attenuation factor; The gradient calculation in backpropagation is expressed as: ； in, is the gradient in back propagation; is the loss function of the neural network, which represents the error between the predicted value and the true value; is the L2 regularization term; is the weight of the neural network, representing the connection strength between the input features of coal mine equipment data and neurons; Under the quantum decoherence suppression mechanism, the gradient calculation is expressed as: ； in, is the gradient after quantum noise suppression, representing the weight update gradient adjusted by the quantum decoherence suppression mechanism; is the quantum noise model, which represents the noise suppression effect; The updating process of coal mine equipment data features is expressed as: ； in, is the updated coal mine equipment data feature vector, representing the coal mine equipment data features after data feature mapping and quantum noise suppression; is the kernel function, representing the input coal mine equipment data points and Projections in high-dimensional space; is the coal mine equipment data feature vector after being mapped by the kernel function, representing the coal mine equipment data feature space after nonlinear transformation; for The corresponding quantum decoherence suppression term.

4. The coal mine equipment fault diagnosis method according to claim 1, characterized in that: Acquire target real-time coal mine equipment data, and use the trained coal mine equipment fault diagnosis model to perform fault diagnosis on the target real-time coal mine equipment data to obtain a final coal mine equipment fault diagnosis result, specifically including: Obtain target real-time coal mine equipment data; Preprocessing the target real-time coal mine equipment data to obtain preprocessed target real-time coal mine equipment data; The pre-processed target real-time coal mining equipment data is input into the trained coal mining equipment fault diagnosis model to obtain a final coal mining equipment fault diagnosis result.

5. The coal mine equipment fault diagnosis method according to claim 4, characterized in that: Preprocessing the target real-time coal mine equipment data to obtain the preprocessed target real-time coal mine equipment data specifically includes: performing data cleaning on the target real-time coal mine equipment data to obtain cleaned target real-time coal mine equipment data; performing data normalization or standardization on the target real-time coal mine equipment data after data cleaning to obtain normalized or standardized target real-time coal mine equipment data; The normalized or standardized target real-time coal mine equipment data is subjected to feature extraction in the time domain and the frequency domain to obtain target real-time coal mine equipment data features as pre-processed target real-time coal mine equipment data.

6. The coal mine equipment fault diagnosis method according to claim 1, characterized in that: After the step of obtaining a trained coal mine equipment fault diagnosis model after iterative training is completed, the coal mine equipment fault diagnosis method further includes: The perturbation-sensing driven model incremental training mechanism is used to continuously perform model incremental training on the trained coal mine equipment fault diagnosis model to obtain a coal mine equipment fault diagnosis model after model incremental training; the coal mine equipment fault diagnosis model after model incremental training is used as the trained coal mine equipment fault diagnosis model to perform fault diagnosis on the target real-time coal mine equipment data to obtain a final coal mine equipment fault diagnosis result.

7. The coal mine equipment fault diagnosis method according to claim 6, characterized in that: The perturbation-sensing driven model incremental training mechanism is used to continuously perform model incremental training on the trained coal mine equipment fault diagnosis model to obtain a coal mine equipment fault diagnosis model after model incremental training, specifically including: Monitor the distribution change of the input coal mine equipment data characteristics in real time, determine whether the distribution change of the coal mine equipment data characteristics meets the characteristic distribution drift trigger condition, and obtain a first judgment result; the characteristic distribution drift trigger condition is: ； in, Indicates the trigger condition for feature distribution drift, is the Kullback-Leibler divergence; is the characteristic distribution of current coal mine equipment data; is the historical distribution during model training; is the preset trigger threshold; When the first judgment result is yes, it is determined that the distribution change of the coal mine equipment data characteristics has caused characteristic distribution drift. At this time, the local perturbation reconstruction mechanism is triggered to perform perturbation incremental update on the affected hidden layer neurons. The weight of each neuron that needs to be updated is expressed as: ； in, is the original neural network connection weight; is the perturbation step length; is the perturbation gradient of the input feature, expressed as: ； in, It is the sliding mean of historical features, used to measure the feature deviation direction; is the regularization factor, used to control the disturbance scale; is a symbolic function; is the loss function of the neural network, which represents the error between the predicted value and the true value; For the coal mine equipment data samples; The memory-preserving regularization term is introduced to update the objective function. The updated objective function is expressed as: ； in, represents the updated objective function; is a snapshot of the key parameters of the original model; is the key parameter of the original neural network model; is the set of parameters that has the greatest impact on performance; is the retention strength factor.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the coal mine equipment fault diagnosis method according to any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the coal mine equipment fault diagnosis method according to any one of claims 1 to 7 is implemented.

Images