Hydraulic system fault diagnosis methods, systems and operating machinery

Abstract

This invention provides a method, system, and hydraulic system for fault diagnosis of hydraulic systems, relating to the field of fault detection technology. The method includes: acquiring operating status data of the hydraulic system; inputting the operating status data into a diagnostic model to obtain fault diagnosis results for the hydraulic system; wherein the diagnostic model is trained based on operating status parameter samples of the hydraulic system under different fault types, and fault type labels corresponding to the operating status parameter samples, and the operating status parameter samples are composed of a combination of multiple operating status data samples acquired from the hydraulic system. This method addresses the shortcomings of existing technologies that predict the fault type of a hydraulic system based on a single type of detection data, resulting in low accuracy in diagnosing hydraulic system faults. It achieves fault diagnosis of hydraulic systems based on multiple types of operating status data, improving the accuracy of the diagnostic results.

Description

Technical Field

[0001] This invention relates to the field of fault diagnosis technology, and in particular to a method, system and working machinery for diagnosing faults in hydraulic systems. Background Technology

[0002] Hydraulic systems are lightweight, powerful, and operate smoothly, making them widely used in mechanical equipment. A hydraulic system consists of a transmission system and a control system, comprising power components, actuators, control components, auxiliary components, and hydraulic oil. The structures of each hydraulic component are complex, and they require the conversion between electrical energy, hydraulic energy, and mechanical energy, resulting in complex operating conditions. Furthermore, the harsh working environment of heavy industrial machinery using hydraulic systems leads to frequent equipment failures, reducing work efficiency and significantly impacting the company's profitability.

[0003] The complex structure and harsh working environment of hydraulic systems make troubleshooting and repair extremely difficult. Traditional fault diagnosis methods require manual feature extraction, selection, and fusion. This process often necessitates extensive and iterative debugging and selection to obtain features that accurately reflect the system's state. Raw information is easily lost or diminished during this complex feature engineering process, increasing the uncertainty of feature extraction and affecting the accuracy of classification results. Furthermore, the computational power of traditional fault diagnosis methods is clearly insufficient to meet the fault diagnosis needs in today's "mechanical big data" context. In this context, deep learning-based fault diagnosis methods have emerged to overcome the shortcomings of traditional methods, enabling adaptive feature extraction and intelligent health diagnosis of mechanical equipment under the trend of big data development.

[0004] However, current deep learning models used for fault diagnosis in hydraulic systems mostly predict faults based on detection data during the operation of the hydraulic system. This is not accurate enough for predicting fault types for hydraulic systems, which have complex structures and operating conditions. Summary of the Invention

[0005] This invention provides a hydraulic system fault diagnosis method, system, and operating machinery to address the shortcomings of existing technologies that predict hydraulic system fault types based on a single type of detection data, resulting in low accuracy of hydraulic system fault diagnosis results. This invention enables the diagnosis of hydraulic system faults based on multiple types of operating status data, thereby improving the accuracy of the diagnosis results.

[0006] This invention provides a method for diagnosing faults in a hydraulic system, comprising:

[0007] Acquire operating status data of the hydraulic system;

[0008] The operating status data is input into the diagnostic model to obtain the fault diagnosis results of the hydraulic system;

[0009] The diagnostic model is trained based on the operating state parameter samples of the hydraulic system under different fault types and the fault type labels corresponding to the operating state parameter samples. The operating state parameter samples are composed of a combination of various operating state data samples obtained from the hydraulic system.

[0010] According to the hydraulic system fault diagnosis method of the present invention, the operating status data is input into the diagnostic model to obtain the fault diagnosis result of the hydraulic system, including:

[0011] The operating status data is subjected to standard normalization processing to obtain operating status parameters;

[0012] Feature extraction is performed on the operating state parameters to obtain the features in the operating state parameters that characterize the working state of the hydraulic system;

[0013] The features are then subjected to dimensionality reduction processing to obtain dimensionality-reduced features;

[0014] The reduced-dimensional features are nonlinearly combined to obtain combined features;

[0015] The combined features are classified to obtain the fault diagnosis results.

[0016] According to the hydraulic system fault diagnosis method of the present invention, the operating status data includes one or more of the following: oil temperature, oil pressure, oil flow rate, vibration frequency, and noise.

[0017] According to the hydraulic system fault diagnosis method of the present invention, the operating status data is subjected to standard normalization processing to obtain operating status parameters, including:

[0018] Based on the types of the operation status data, the operation status data is classified to obtain various types of operation status data divided by type;

[0019] Based on the total amount of data for each type of operating status, the mean value of each type of operating status data is calculated.

[0020] Based on the various types of operational status data, the total amount of data, and the mean, the standard deviation of the various types of operational status data is calculated.

[0021] Based on the various types of operating status data, the mean, and the standard deviation, the operating status parameters of the various types of operating status data are obtained;

[0022] The operating status parameters of the various types of operating status data are combined to obtain the operating status parameters obtained after the operating status data has been normalized according to standards.

[0023] According to the hydraulic system fault diagnosis method of the present invention, the diagnostic model is constructed as follows:

[0024] A preliminary diagnostic model is trained based on the operating status parameter samples and the fault type labels corresponding to the operating status parameter samples;

[0025] The hyperparameters of the pre-diagnostic model were tuned.

[0026] The preliminary diagnostic model, after hyperparameter tuning, is used as the diagnostic model.

[0027] According to the hydraulic system fault diagnosis method of the present invention, hyperparameter tuning of the pre-diagnostic model includes:

[0028] Based on the fault type prediction result obtained by inputting the operating status parameter sample into the pre-diagnostic model, and the fault type label corresponding to the operating status parameter sample, an objective function is obtained to characterize the error between the fault type prediction result and the fault type label.

[0029] Based on the objective function and the hyperparameters at the previous time step, the gradient of the objective function at the current time step is obtained;

[0030] Based on the gradient of the objective function and the expectation at the previous time step, the expectation of the gradient of the objective function at the current time step is obtained;

[0031] Based on the expectation of the gradient of the objective function at the current time step, the bias correction at the current time step is obtained;

[0032] Based on the bias correction at the current time step and the hyperparameters at the previous time step, the hyperparameters at the current time step are obtained.

[0033] Based on the hyperparameters at the current time step, update the hyperparameters at the previous time step until the objective function converges.

[0034] The present invention also provides a hydraulic system fault diagnosis system, comprising:

[0035] The acquisition module is used to acquire the operating status data of the hydraulic system;

[0036] The processing module is used to input the operating status data into the diagnostic model to obtain the fault diagnosis results of the hydraulic system;

[0037] The diagnostic model is trained based on the operating state parameter samples of the hydraulic system under different fault types and the fault type labels corresponding to the operating state parameter samples. The operating state parameter samples are composed of a combination of various operating state data samples obtained from the hydraulic system.

[0038] The present invention also provides a working machine including the hydraulic system fault diagnosis system as described above.

[0039] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the hydraulic system fault diagnosis method as described above.

[0040] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the hydraulic system fault diagnosis method as described above.

[0041] This invention provides a hydraulic system fault diagnosis method, system, and operating machinery. By inputting the acquired hydraulic system operating status data into a diagnostic model pre-trained based on operating status parameter samples of the hydraulic system under different fault types and fault type labels corresponding to the operating status parameter samples, the fault diagnosis result of the hydraulic system is obtained. At the same time, the operating status parameter samples are composed of a combination of various operating status data samples acquired from the hydraulic system, enabling the diagnostic model to make comprehensive predictions based on multiple factors causing hydraulic system faults, effectively improving the accuracy of judging the fault type of the hydraulic system based on the diagnostic model.

[0042] By standardizing and normalizing the operational status data containing multiple data types, the features of each dimension are normalized to the same value range, eliminating the correlation between different features. This allows the diagnostic model to comprehensively analyze features of different dimensions, ensuring the accuracy of fault type prediction.

[0043] A preliminary diagnostic model is obtained by training based on samples of operating status parameters and corresponding fault type labels. Then, the hyperparameters of the preliminary diagnostic model are optimized, and the optimized preliminary diagnostic model is used as the diagnostic model, thereby further improving the accuracy of hydraulic system fault diagnosis. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0045] Figure 1 This is a flowchart illustrating a hydraulic system fault diagnosis method provided by the present invention;

[0046] Figure 2 This is a schematic diagram of the structure of a hydraulic system fault diagnosis system provided by the present invention;

[0047] Figure 3 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0049] It should be noted that traditional fault diagnosis methods for hydraulic systems require manual feature extraction, selection, and fusion. This process often necessitates extensive and iterative debugging and selection to obtain features that accurately reflect the system's state. Raw information is easily diminished or lost during this complex feature engineering process, increasing the uncertainty of feature extraction and affecting the accuracy of classification results. Furthermore, the computational power of traditional fault diagnosis methods is clearly insufficient to meet the fault diagnosis needs in today's "mechanical big data" context. In this context, deep learning-based fault diagnosis methods have emerged to overcome the shortcomings of traditional methods, enabling adaptive feature extraction and intelligent health diagnosis of mechanical equipment under the trend of big data development.

[0050] However, current deep learning-based fault diagnosis methods often only predict fault types based on one type of detection data for hydraulic systems. However, hydraulic systems have complex structures, harsh working environments, and complex operating conditions. Faults can be categorized into at least three main types: abnormal pressure, abnormal motion, and abnormal speed. These faults can manifest in various ways, including abnormal temperature, vibration, noise, and abnormal oil pressure. Therefore, fault type prediction based on a single detection data point obviously cannot guarantee the accuracy of fault diagnosis for hydraulic systems.

[0051] Based on this, this invention proposes a hydraulic system fault diagnosis method. By inputting the acquired hydraulic system operating status data into an operating status parameter sample composed of multiple operating status data samples acquired from the hydraulic system under different fault types, and a diagnostic model trained with fault type labels corresponding to the operating status parameter samples, the fault diagnosis result is obtained. This method makes the fault diagnosis of the hydraulic system based on multiple monitoring data of the hydraulic system during operation, effectively improving the accuracy of fault diagnosis.

[0052] The following is combined Figure 1 This invention describes a hydraulic system fault diagnosis method, executed by the controller of the hydraulic system or its software and / or hardware, such as... Figure 1 As shown, the method includes the following steps:

[0053] 101. Obtain the operating status data of the hydraulic system;

[0054] Specifically, during operation, the temperature, oil pressure, and flow rate of a hydraulic system will all change. When a fault occurs, it will manifest in certain aspects of the hydraulic system. For example, when air gets into the hydraulic system's suction circuit, or when oil becomes trapped, or when vanes or pistons seize, abnormal vibration and noise will occur. Poor heat dissipation, low oil level in the tank, or air or moisture mixed in the oil will cause the hydraulic system to overheat. Blockage of the pressure reducing valve, relief valve, or unloading valve, or malfunction of the variable valve mechanism, will cause excessively high oil pressure. Damage to the relief valve, bypass valve, or hydraulic valve will result in insufficient oil pressure. Therefore, by accurately analyzing the monitored operating data of the hydraulic system, such as temperature, oil pressure, vibration, and noise, it is possible to determine whether a fault has occurred in the hydraulic system and to identify the type of fault.

[0055] 102. Input the operating status data into the diagnostic model to obtain the fault diagnosis results of the hydraulic system;

[0056] The diagnostic model is trained based on the operating state parameter samples of the hydraulic system under different fault types and the fault type labels corresponding to the operating state parameter samples. The operating state parameter samples are composed of a combination of various operating state data samples obtained from the hydraulic system.

[0057] Specifically, by collecting operating status data of the hydraulic system under different fault types, and then combining the operating status data to form operating status parameters, a diagnostic model for hydraulic system fault diagnosis can be obtained by training the operating status parameter samples and corresponding fault type labels, thereby realizing real-time diagnosis of hydraulic system faults.

[0058] More specifically, by combining multiple operating state data samples to form operating state parameter samples, and then using the operating state parameter samples and labels to train a diagnostic model, the diagnostic model can make a comprehensive judgment on the operating state of the hydraulic system based on multiple operating state data during the operation of the hydraulic system, thereby accurately predicting the fault type of the hydraulic system, facilitating subsequent targeted maintenance of the hydraulic system, effectively improving the operating efficiency of the hydraulic system, and ensuring the progress of the project.

[0059] As an embodiment of the present invention, the operating status data is input into the diagnostic model to obtain the fault diagnosis result of the hydraulic system, including:

[0060] The operating status data is subjected to standard normalization processing to obtain the operating status parameters;

[0061] Feature extraction is performed on the operating state parameters to obtain the features in the operating state parameters that characterize the working state of the hydraulic system;

[0062] The features are then subjected to dimensionality reduction processing to obtain dimensionality-reduced features;

[0063] The reduced-dimensional features are nonlinearly combined to obtain combined features;

[0064] The combined features are classified to obtain the fault diagnosis results.

[0065] Specifically, the types of data that can reflect hydraulic system faults are divided into several categories, such as oil pressure, temperature, vibration, and noise. When these operating status data are input into the diagnostic model, they are regarded as features in different dimensions according to their different types. It is understandable that the source and measurement unit of each feature are generally different, so the distribution range of feature values ​​also varies greatly. By standardizing the operating status data, the features of each dimension are normalized to the same value range, eliminating the correlation between different features. This allows the diagnostic model to comprehensively analyze the features of different dimensions, ensuring the accuracy of fault type prediction.

[0066] More specifically, the diagnostic model extracts features from the operating state parameters to identify the types that can characterize the hydraulic system faults. Then, it performs dimensionality reduction on the features, i.e., downsampling, to remove the secondary parts of the obtained feature maps, thereby avoiding overfitting. After that, the dimensionality-reduced features are nonlinearly combined to obtain the output. Finally, the previously described network layer combination is used to convert the input operating state data into the final category, i.e., to obtain the fault diagnosis result.

[0067] It's understandable that the process of diagnosing hydraulic system faults based on operational status data using a diagnostic model is consistent with the training process of that model. In other words, during the training of the diagnostic model, when operational status data samples are input into the initial model, the initial model also performs standard normalization on the samples. This is because different feature value ranges in the operational status data samples can cause gradient directions at most locations to be non-optimal search directions. This would require multiple iterations of the network to converge during training. Therefore, data normalization is necessary to normalize the features across different dimensions to the same value range, eliminating correlations between different features and improving training efficiency. Furthermore, since different feature indicators have different dimensions and units, without normalization, some indicators might be overlooked, affecting the data analysis results. Subsequently, after obtaining the operating status parameter samples, the model undergoes feature extraction, feature dimensionality reduction, nonlinear combination, and classification in sequence. Finally, the model parameters are adjusted based on the fault diagnosis results obtained from the diagnostic model and the actual labels corresponding to the operating status data samples, until the fault diagnosis results and labels obtained from the diagnostic model are compatible, which indicates that the diagnostic model training is complete.

[0068] As an embodiment of the present invention, the operating status data includes one or more of the following: oil temperature, oil pressure, oil flow rate, vibration frequency, and noise.

[0069] Specifically, as mentioned above, different types of hydraulic system faults are reflected in different operating parameters of the hydraulic system. Therefore, when diagnosing hydraulic system faults, simultaneously acquiring one or more of the following parameters can improve the accuracy of diagnosing hydraulic system fault types: hydraulic oil temperature, hydraulic oil pressure, hydraulic oil flow rate, vibration frequency, and noise.

[0070] More specifically, the data types of operating status described in the embodiments of the present invention are only relatively common data types that can reflect hydraulic system faults and are exemplary. Other data types of operating status not listed in the embodiments of the present invention are also within the protection scope of the technical solution of the present invention.

[0071] As an embodiment of the present invention, the operating status data is subjected to standard normalization processing to obtain operating status parameters, including:

[0072] Based on the types of the operation status data, the operation status data is classified to obtain various types of operation status data divided by type;

[0073] Based on the total amount of data for each type of operating status, the mean value of each type of operating status data is calculated.

[0074] Based on the various types of operational status data, the total amount of data, and the mean, the standard deviation of the various types of operational status data is calculated.

[0075] Based on the various types of operating status data, the mean, and the standard deviation, the operating status parameters of the various types of operating status data are obtained;

[0076] The operating status parameters of the various types of operating status data are combined to obtain the operating status parameters obtained after the operating status data has been normalized according to standards.

[0077] Specifically, normalization of data can improve the training efficiency of the model and enable the model to achieve higher accuracy. Therefore, after inputting the running status data into the diagnostic model, standard normalization is first performed to obtain the running status parameters.

[0078] More specifically, regarding the standard normalization process used in the embodiments of the present invention, with N samples... For example, for each feature x, its mean and standard deviation are as follows:

[0079]

[0080]

[0081] Based on the mean and standard deviation, the operating status parameters for various types of operating status data can be obtained as shown in Equation 3:

[0082]

[0083] As one embodiment of the present invention, the diagnostic model is constructed as follows:

[0084] A preliminary diagnostic model is trained based on the operating status parameter samples and the fault type labels corresponding to the operating status parameter samples;

[0085] The hyperparameters of the pre-diagnostic model were tuned.

[0086] The preliminary diagnostic model, after hyperparameter tuning, is used as the diagnostic model.

[0087] Understandably, during the training of a convolutional neural network model, the traditional stochastic gradient descent method updates the weights with a constant learning rate. Therefore, when the learning rate is not set properly, the training may get stuck in a local minimum or a saddle point, affecting the performance of the entire model.

[0088] Based on this, in the above embodiments of the present invention, after training a preliminary diagnostic model based on the operating state parameter samples and the fault type labels corresponding to the operating state parameter samples, the preliminary diagnostic model is subjected to hyperparameter tuning, and then the preliminary diagnostic model after hyperparameter tuning is used as the diagnostic model. That is, the hyperparameter learning rate of the preliminary diagnostic model is periodically adjusted to improve the accuracy of the obtained diagnostic model in diagnosing hydraulic system faults.

[0089] As an embodiment of the present invention, hyperparameter tuning of the pre-diagnostic model includes:

[0090] Based on the fault type prediction result obtained by inputting the operating status parameter sample into the pre-diagnostic model, and the fault type label corresponding to the operating status parameter sample, an objective function is obtained to characterize the error between the fault type prediction result and the fault type label.

[0091] Based on the objective function and the hyperparameters at the previous time step, the gradient of the objective function at the current time step is obtained;

[0092] Based on the gradient of the objective function and the expectation at the previous time step, the expectation of the gradient of the objective function at the current time step is obtained;

[0093] Based on the expectation of the gradient of the objective function at the current time step, the bias correction at the current time step is obtained;

[0094] Based on the bias correction at the current time step and the hyperparameters at the previous time step, the hyperparameters at the current time step are obtained.

[0095] Based on the hyperparameters at the current time step, update the hyperparameters at the previous time step until the objective function converges.

[0096] Specifically, during the training process of the diagnostic model, when the prediction result output based on the running state data sample of the input model does not meet the expected value, the parameters of the model are adjusted through the backpropagation process. The backpropagation process is the process of hyperparameter tuning of the pre-diagnostic model.

[0097] More specifically, the operational status data samples are passed from the input layer through convolutional layers, pooling layers, and fully connected layers to the output layer. Data loss occurs during this transfer between layers, leading to errors. The error between the output result and the label (the expected value) is calculated, and this error is input to the output. The error is then recursively calculated layer by layer using backpropagation. The Adam algorithm (Adaptive Moment Estimation) is used to adjust the connection weights of each layer; this constitutes one backpropagation process. This process is repeated until the error between the output value and the expected value is reduced to a certain value, at which point the weights stop updating, and the training of the diagnostic model is complete.

[0098] Understandably, the main purpose of backpropagation is to adjust the network weights using training samples and expected values. Since the hydraulic system fault diagnosis problem is a multi-class classification problem, the minimum cross-entropy loss function is used to calculate the error between the output and the true value.

[0099]

[0100] Where y is the output value, i.e. the predicted value of the sample; n represents the total number of samples; and a is the expected value, i.e. the label.

[0101] Furthermore, the Adam optimization algorithm is used to update the weights of the preliminary diagnostic model. The specific steps include:

[0102]

[0103] m t =μ*m t-1 +(1-μ)*g t (6)

[0104]

[0105] Among them, g t Let m be the gradient of the objective function. t and n t Let μ and v be the first and second moment estimates of the gradient, respectively, and let μ and v be the exponential decay rates of the first and second moment estimates of the gradient, respectively, and μ, v∈[0,1].

[0106] Then at the t-th iteration:

[0107]

[0108]

[0109] Therefore, the expression for updating the parameters of the preliminary diagnostic model is:

[0110]

[0111] Where α is the learning rate, and ε is a very small number to prevent the denominator from being zero. θ t These are the updated parameters, including weights and biases.

[0112] The hydraulic system fault diagnosis method described in the above embodiments of the present invention adaptively adjusts the parameter learning rate by calculating the first-order moment estimate and second-order moment estimate of the gradient, and updates the neural network weights by iteratively updating the training data. Simultaneously, during the optimization process in gradient descent, the parameter θ is updated. t The learning rate is adjusted to generate adaptive learning rates for different parameters, thereby avoiding the problem of getting stuck in local minima or saddle points when training diagnostic models based on stochastic gradient descent.

[0113] Furthermore, the hydraulic system fault diagnosis method described in this embodiment of the invention estimates the type of hydraulic system fault based on operating status data including oil temperature, oil pressure, oil flow rate, vibration frequency, and noise. After standard normalization, each operating status data forms a one-dimensional time series signal, which enables the convolutional layer used for feature extraction in the diagnostic model to be a one-dimensional convolutional layer sequence. At the same time, the number of convolutional kernels does not need to exceed 10, which greatly reduces the training difficulty of the diagnostic model and improves the efficiency of using the diagnostic model for hydraulic system fault diagnosis.

[0114] The present invention provides a hydraulic system fault diagnosis system. The hydraulic system fault diagnosis system described below and the hydraulic system fault diagnosis method described above can be referred to in correspondence.

[0115] like Figure 2 As shown, a hydraulic system fault diagnosis system provided in an embodiment of the present invention includes: an acquisition module 210 and a processing module 220; wherein,

[0116] The acquisition module 210 is used to acquire the operating status data of the hydraulic system;

[0117] The processing module 220 is used to input the operating status data into the diagnostic model to obtain the fault diagnosis result of the hydraulic system.

[0118] The diagnostic model is trained based on the operating state parameter samples of the hydraulic system under different fault types and the fault type labels corresponding to the operating state parameter samples. The operating state parameter samples are composed of a combination of various operating state data samples obtained from the hydraulic system.

[0119] The hydraulic system fault diagnosis system described in this embodiment of the invention obtains the fault diagnosis result of the hydraulic system by inputting the acquired hydraulic system operating status data into a diagnostic model pre-trained based on operating status parameter samples of the hydraulic system under different fault types and fault type labels corresponding to the operating status parameter samples. At the same time, the operating status parameter samples are composed of a combination of various operating status data samples acquired from the hydraulic system, enabling the diagnostic model to make comprehensive predictions based on multiple factors causing hydraulic system faults, effectively improving the accuracy of judging the fault type of the hydraulic system based on the diagnostic model.

[0120] Preferably, the processing module is specifically used to perform standard normalization processing on the operating status data to obtain the operating status parameters; to extract features from the operating status parameters to obtain features that characterize the working state of the hydraulic system; to perform dimensionality reduction processing on the features to obtain dimensionality-reduced features; to perform nonlinear combination of the dimensionality-reduced features to obtain combined features; and to classify the combined features to obtain the fault diagnosis result.

[0121] Preferably, the operating status data acquired by the acquisition module includes one or more of the following: oil temperature, oil pressure, oil flow rate, vibration frequency, and noise.

[0122] Preferably, the processing module is further configured to classify the operating status data based on the type of the operating status data to obtain various types of operating status data; calculate the mean of each type of operating status data based on the total amount of data in each type of operating status data; calculate the standard deviation of each type of operating status data based on the various types of operating status data, the total amount of data, and the mean; obtain the operating status parameters of each type of operating status data based on the various types of operating status data, the mean, and the standard deviation; and combine the operating status parameters of each type of operating status data to obtain the operating status parameters obtained after standard normalization processing of the operating status data.

[0123] Preferably, the diagnostic model is constructed as follows:

[0124] A preliminary diagnostic model is trained based on the operating status parameter samples and the fault type labels corresponding to the operating status parameter samples;

[0125] The hyperparameters of the pre-diagnostic model were tuned.

[0126] The preliminary diagnostic model, after hyperparameter tuning, is used as the diagnostic model.

[0127] Preferably, hyperparameter tuning of the pre-diagnostic model includes:

[0128] Based on the fault type prediction result obtained by inputting the operating state parameter samples into the pre-diagnostic model, and the fault type label corresponding to the operating state parameter samples, an objective function characterizing the error between the fault type prediction result and the fault type label is obtained; based on the objective function and the hyperparameters at the previous time step, the gradient of the objective function at the current time step is obtained; based on the gradient of the objective function and the expectation at the previous time step, the expectation of the gradient of the objective function at the current time step is obtained; based on the expectation of the gradient of the objective function at the current time step, the bias correction at the current time step is obtained; based on the bias correction at the current time step and the hyperparameters at the previous time step, the hyperparameters at the current time step are obtained; and based on the hyperparameters at the current time step, the hyperparameters at the previous time step are updated until the objective function converges.

[0129] This invention also provides a working machine that includes a hydraulic system fault diagnosis system as described in any of the above embodiments.

[0130] Specifically, the working machinery that includes the hydraulic system fault diagnosis system described in this invention has all the advantages and technical effects of the hydraulic system fault diagnosis system, which will not be repeated here.

[0131] Figure 3 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 3 As shown, the electronic device may include a processor 310, a communication interface 320, a memory 330, and a communication bus 340, wherein the processor 310, the communication interface 320, and the memory 330 communicate with each other via the communication bus 340. The processor 310 can call logical instructions in the memory 330 to execute a hydraulic system fault diagnosis method. This method includes: acquiring operating status data of the hydraulic system; inputting the operating status data into a diagnostic model to obtain a fault diagnosis result for the hydraulic system; wherein the diagnostic model is trained based on operating status parameter samples of the hydraulic system under different fault types, and fault type labels corresponding to the operating status parameter samples, and the operating status parameter samples are composed of a combination of various operating status data samples acquired from the hydraulic system.

[0132] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0133] On the other hand, the present invention also provides a computer program product, the computer program product including a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, when the program instructions are executed by a computer, the computer is able to execute a hydraulic system fault diagnosis method provided by the above methods, the method including: acquiring operating status data of the hydraulic system; inputting the operating status data into a diagnostic model to obtain a fault diagnosis result of the hydraulic system; wherein, the diagnostic model is trained based on operating status parameter samples of the hydraulic system under different fault types and fault type labels corresponding to the operating status parameter samples, the operating status parameter samples being composed of a combination of multiple operating status data samples acquired from the hydraulic system.

[0134] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program is capable of implementing a hydraulic system fault diagnosis method provided by the above methods. The method includes: acquiring operating status data of the hydraulic system; inputting the operating status data into a diagnostic model to obtain a fault diagnosis result of the hydraulic system; wherein the diagnostic model is trained based on operating status parameter samples of the hydraulic system under different fault types and fault type labels corresponding to the operating status parameter samples, and the operating status parameter samples are composed of a combination of multiple operating status data samples acquired from the hydraulic system.

[0135] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0136] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0137] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A hydraulic system failure diagnosis method characterized by, The method comprises the following steps: acquiring operation state data of a hydraulic system; inputting the operation state data into a diagnosis model to obtain a fault diagnosis result of the hydraulic system; wherein the diagnosis model is trained based on operation state parameter samples of the hydraulic system under different fault types and fault type labels corresponding to the operation state parameter samples, and the operation state parameter samples are composed of multiple operation state data samples obtained from the hydraulic system; the diagnosis model is constructed in the following manner: a preliminary diagnosis model is trained based on the operation state parameter samples and the fault type labels corresponding to the operation state parameter samples; the preliminary diagnosis model is subjected to hyperparameter tuning; the preliminary diagnosis model subjected to the hyperparameter tuning is used as the diagnosis model; the hyperparameter tuning of the preliminary diagnosis model comprises the following steps: a target function representing the error between a fault type prediction result obtained by inputting the operation state parameter samples into the preliminary diagnosis model and the fault type labels corresponding to the operation state parameter samples is obtained; a gradient of the target function at a current time step is obtained based on the target function and the hyperparameters at a previous time step; an expectation of the gradient of the target function at the current time step is obtained based on the gradient of the target function and an expectation at the previous time step; a bias correction at the current time step is obtained based on the expectation of the gradient of the target function at the current time step; the hyperparameters at the current time step are obtained based on the bias correction at the current time step and the hyperparameters at the previous time step; the hyperparameters at the previous time step are updated based on the hyperparameters at the current time step until the target function converges.

2. The hydraulic system failure diagnosis method according to claim 1, characterized by, The operation state data is inputted into the diagnosis model to obtain the fault diagnosis result of the hydraulic system, which comprises the following steps: the operation state data is subjected to standard normalization processing to obtain operation state parameters; features representing the working state of the hydraulic system in the operation state parameters are extracted to obtain the operation state parameters; the features are subjected to dimension reduction processing to obtain reduced dimension features; the reduced dimension features are subjected to nonlinear combination to obtain combined features; the combined features are subjected to classification to obtain the fault diagnosis result.

3. The hydraulic system failure diagnosis method according to claim 2, characterized by, The operation state data comprises one or more of oil temperature, oil pressure, oil flow, vibration frequency and noise.

4. The hydraulic system failure diagnosis method according to claim 3, characterized by, The operation state data is subjected to standard normalization processing to obtain operation state parameters, which comprises the following steps: the operation state data is classified based on the types of the operation state data to obtain operation state data of each type classified by type; the mean value of the operation state data of each type is calculated based on the total amount of data of the operation state data of each type; the standard deviation of the operation state data of each type is calculated based on the operation state data of each type, the total amount of data and the mean value. Obtaining running state parameters of the various types of running state data based on the various types of running state data, the mean values, and the standard deviations; Combining the running state parameters of the various types of running state data to obtain the running state parameters obtained after the running state data is processed by standard normalization.

5. A hydraulic system failure diagnosis system characterized by comprising: The hydraulic system fault diagnosis system comprises: An acquisition module configured to acquire running state data of a hydraulic system; A processing module configured to input the running state data into a diagnosis model to obtain a fault diagnosis result of the hydraulic system; The diagnosis model is trained based on running state parameter samples of the hydraulic system under different fault types and fault type labels corresponding to the running state parameter samples, and the running state parameter samples are composed of multiple running state data samples acquired from the hydraulic system; The diagnosis model is constructed in the following manner: A preliminary diagnosis model is trained based on the running state parameter samples and the fault type labels corresponding to the running state parameter samples; The preliminary diagnosis model is hyperparameter-optimized; The preliminary diagnosis model that has been hyperparameter-optimized is used as the diagnosis model; The hyperparameter-optimization of the preliminary diagnosis model comprises: A target function representing errors between fault type prediction results obtained by inputting the running state parameter samples into the preliminary diagnosis model and fault type labels corresponding to the running state parameter samples is obtained; The gradient of the target function at the current time step is obtained based on the target function and the hyperparameters at the previous time step; The expectation of the gradient of the target function at the current time step is obtained based on the gradient of the target function and the expectation at the previous time step; The bias correction at the current time step is obtained based on the expectation of the gradient of the target function at the current time step; The hyperparameters at the current time step are obtained based on the bias correction at the current time step and the hyperparameters at the previous time step; The hyperparameters at the previous time step are updated based on the hyperparameters at the current time step until the target function converges.

6. A work machine characterized by, The hydraulic system fault diagnosis system of claim 5 is included.

7. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor implements the hydraulic system fault diagnosis method of any one of claims 1 to 4 when executing the program.

8. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program implements the hydraulic system fault diagnosis method of any one of claims 1 to 4 when executed by the processor.

Images