Stator winding end vibration fault diagnosis model acquisition method and system

Abstract

The invention relates to a stator winding end vibration fault diagnosis model obtaining method and system.The stator winding end vibration fault diagnosis model obtaining method comprises the steps that historical operation data are obtained, the historical operation data comprise operation parameters and stator winding end amplitudes at corresponding moments, and the operation parameters are calculated through Kendall correlation coefficients; determining a target parameter from the operation parameters, and determining training data according to the target parameter and the corresponding stator winding end amplitude; iteratively training a radial basis function generalized regression model based on the training data and a gradient descent method to obtain a theoretical value of the amplitude of the end part of the stator winding, and calculating an error between the theoretical value and a true value; and when the error is smaller than a first threshold value or the number of iterations is equal to a preset number of iterations, iteration is stopped, and a stator winding end vibration fault diagnosis model is obtained.

Description

Technical Field

[0001] The present application relates to the field of large generator fault diagnosis, and in particular to a method and system for acquiring a stator winding end vibration fault diagnosis model. Background Art

[0002] As a traditional core power generation equipment, the safe and efficient operation of large steam turbine generators directly affects the stability of the power system. Among them, the stator winding plays a vital role in converting mechanical energy into electrical energy. The vibration of the stator winding end is one of the important parameters that need to be monitored and managed during the operation of the steam turbine generator. Under the action of long-term electromagnetic force, thermal stress, and mechanical stress, it is inevitable that the stator winding end will become loose, causing abnormal vibration of the stator winding end. If the vibration fault is not discovered and handled in time, it may cause equipment damage or accidents, and even endanger the stable operation of the power system.

[0003] At present, the diagnosis of generator stator end vibration faults is generally done by setting a fixed threshold range and issuing an alarm when the threshold range is exceeded. However, this method has obvious limitations. In the early stage of a fault, the fixed threshold rule cannot effectively deal with the situation where the vibration amplitude is weak but the vibration may indicate a potential problem, resulting in delayed or missed detection of potential problems, thereby increasing the risk of unexpected downtime or equipment damage.

[0004] The existing generator stator end vibration fault diagnosis method has problems of poor real-time performance and poor diagnostic effect. Summary of the invention

[0005] The embodiments of the present application provide a method and system for obtaining a stator winding end vibration fault diagnosis model, so as to at least solve the problems of poor real-time performance and poor diagnostic effect in the related art.

[0006] In a first aspect, an embodiment of the present application provides a method for obtaining a stator winding end vibration fault diagnosis model, comprising:

[0007] Acquire historical operation data, the historical operation data including operation parameters and stator winding end amplitudes at corresponding moments, determine target parameters from the operation parameters by using Kendall correlation coefficients, and determine training data according to the target parameters and the corresponding stator winding end amplitudes;

[0008] Iteratively training a radial basis function generalized regression model based on the training data and the gradient descent method to obtain a theoretical value of the stator winding end amplitude, and calculating an error between the theoretical value and the true value;

[0009] In response to the error being less than a first threshold or the number of iterations being equal to a preset number of iterations, the iteration is stopped to obtain a stator winding end vibration fault diagnosis model.

[0010] In one embodiment, the radial basis function generalized regression model is configured to use a Gaussian permutation function as a kernel function.

[0011] In one embodiment, the radial basis function generalized regression model is configured to: cluster the features of the training data through self-organizing mapping, and determine the number of sets, the center point and radius of each set.

[0012] In one embodiment, the radial basis function generalized regression model includes an input layer, a hidden layer and an output layer, wherein the hidden layer is configured to perform spatial mapping using a Gaussian permutation function as a kernel function, and the Gaussian permutation function is expressed as:

[0013]

[0014] Among them, x i represents the i-th data in the training data, c j represents the center of the jth set, δ j represents the diffusion coefficient of the jth set.

[0015] In one embodiment, the Kendall correlation coefficient represents the correlation between the operating parameter and the amplitude, and the calculation formula of the Kendall correlation coefficient is as follows:

[0016]

[0017] Wherein, τ represents the Kendall correlation coefficient, N is the number of the historical operation data, C represents the number of element logarithms of each operation parameter being positively correlated with the change pattern of the amplitude, and D represents the number of element logarithms of each operation parameter being negatively correlated with the change pattern of the amplitude.

[0018] In one embodiment, the operating parameters include at least two types, and determining the target parameter from the operating parameters includes: selecting the operating parameter whose absolute value of the Kendall correlation coefficient is greater than a second threshold as the target parameter.

[0019] In a second aspect, an embodiment of the present application provides a method for diagnosing a stator winding end vibration fault, comprising:

[0020] Acquire the real-time operating parameters of the steam turbine generator, and obtain the theoretical value of the amplitude of the stator winding end according to the real-time operating parameters through the stator winding end vibration fault diagnosis model described in the first aspect;

[0021] The deviation between the actual value of the amplitude monitored by the sensor and the theoretical value is calculated, and whether there is abnormality in the vibration of the stator winding end is determined based on the deviation.

[0022] In one embodiment, judging whether the vibration of the stator winding end is abnormal according to the deviation includes:

[0023] A ratio of the deviation to the theoretical value is calculated, and in response to the ratio being greater than a third threshold, it is determined that an abnormality exists.

[0024] In a third aspect, an embodiment of the present application provides a stator winding end vibration fault diagnosis model acquisition system, comprising:

[0025] Acquisition module: used to acquire historical operation data, the historical operation data includes operation parameters and stator winding end amplitude at corresponding moments, determine target parameters from the operation parameters through Kendall correlation coefficient, and determine training data according to the target parameters and the corresponding stator winding end amplitude;

[0026] Training module: used for iteratively training the radial basis function generalized regression model based on the training data and the gradient descent method, obtaining the theoretical value of the amplitude of the stator winding end, and calculating the error between the theoretical value and the true value;

[0027] Model module: used to stop iteration in response to the error being less than a first threshold or the number of iterations being equal to a preset number of iterations, so as to obtain a stator winding end vibration fault diagnosis model.

[0028] In a fourth aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for acquiring a stator winding end vibration fault diagnosis model as described in the first aspect above is implemented.

[0029] The method and system for obtaining a stator winding end vibration fault diagnosis model provided in the embodiments of the present application have at least the following technical effects.

[0030] The present application establishes a dynamic stator winding end vibration fault diagnosis model through a radial basis function generalized regression model. Through the Kendall correlation coefficient, the parameters with a high correlation with the stator winding end vibration are determined from the operating parameters, and redundant data are filtered out, thereby improving the working efficiency of the model. Compared with the traditional monitoring method that relies on sensors to collect vibration data and perform fixed threshold alarms, the diagnostic model provided by the present application can more accurately predict the theoretical amplitude of the stator winding end under different working conditions and achieve early fault warning. It helps to intervene and repair the problem before it expands, reduces the risk of unexpected downtime, and improves the accuracy and reliability of fault detection and the safety and stability of motor operation.

[0031] Details of one or more embodiments of the present application are set forth in the following drawings and description to make other features, objects, and advantages of the present application more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0033] Figure 1 is a flow chart of a method for obtaining a stator winding end vibration fault diagnosis model according to an embodiment of the present application;

[0034] Figure 2 is a flow chart of a stator winding end vibration fault diagnosis method provided according to an embodiment of the present application;

[0035] Figure 3 It is a structural block diagram of a stator winding end vibration fault diagnosis model acquisition system provided according to an embodiment of the present application;

[0036] Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0037] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. Based on the embodiments provided in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.

[0038] Obviously, the drawings described below are only some examples or embodiments of the present application. For ordinary technicians in this field, the present application can also be applied to other similar scenarios based on these drawings without creative work. In addition, it can also be understood that although the efforts made in this development process may be complicated and lengthy, for ordinary technicians in this field related to the content disclosed in this application, some changes in design, manufacturing or production based on the technical content disclosed in this application are just conventional technical means, and should not be understood as insufficient content disclosed in this application.

[0039] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those of ordinary skill in the art that the embodiments described in this application may be combined with other embodiments without conflict.

[0040] Unless otherwise defined, the technical terms or scientific terms involved in this application should be understood by people with ordinary skills in the technical field to which this application belongs. The words "one", "a", "a", "the" and the like involved in this application do not indicate a quantitative limitation, and may represent the singular or plural. The terms "include", "comprise", "have" and any of their variations involved in this application are intended to cover non-exclusive inclusions; for example, a process, method, system, product or device that includes a series of steps or modules (units) is not limited to the listed steps or units, but may also include steps or units that are not listed, or may also include other steps or units inherent to these processes, methods, products or devices. The words "connect", "connected", "coupled" and the like involved in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The "multiple" involved in this application refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there may be three relationships, for example, "A and / or B" can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the objects before and after are in an "or" relationship. The terms "first", "second", "third", etc. involved in this application are only used to distinguish similar objects and do not represent a specific ordering of the objects.

[0041] In a first aspect, an embodiment of the present application provides a method for obtaining a stator winding end vibration fault diagnosis model. Figure 1 is a flow chart of a method for obtaining a stator winding end vibration fault diagnosis model according to an embodiment of the present application, such as Figure 1 As shown, the method includes:

[0042] Step S101, obtaining historical operation data, the historical operation data including operation parameters and stator winding end amplitude at corresponding moments, determining target parameters from the operation parameters through Kendall correlation coefficient, and determining training data according to the target parameters and the corresponding stator winding end amplitude.

[0043] Optionally, the operating parameters include stator current, stator voltage, active power, reactive power, load change rate, cooling water inlet temperature and flow rate, hydrogen pressure, hydrogen temperature, hydrogen flow rate, oil temperature and other parameters. According to the kendall correlation coefficient, a target parameter with a high correlation with the stator winding end amplitude at the corresponding moment is selected from multiple operating parameters as training data. In this way, the operating parameters in the historical data are filtered, the interference data with low correlation is removed, and the quality of the training data is improved, which is conducive to improving the prediction accuracy of the subsequent stator winding end vibration fault diagnosis model.

[0044] In one example, the Kendall correlation coefficient characterizes the correlation between the operating parameter and the amplitude. The calculation formula of the Kendall correlation coefficient is as follows:

[0045]

[0046] Among them, τ represents the Kendall correlation coefficient, N is the number of historical operation data, C represents the number of element logarithms of each operating parameter and the change law of amplitude, and D represents the number of element logarithms of each operating parameter and the change law of amplitude.

[0047] Alternatively, taking the stator current as an example, the stator current in the historical operation data is represented by x, and the corresponding stator winding end amplitude is represented by y. The stator current and the corresponding amplitude are a pair of data, and N is the number of data logs in the historical operation data. C means that the stator current and the corresponding amplitude satisfy (x m -x n )(y m -y n )≥0, D means the stator current and the corresponding amplitude satisfy (x m -x n )(y m -y n )<0. m 、x n Respectively represent the mth and nth stator currents, y m ,y n Represent the mth and nth amplitudes respectively.

[0048] Among them, the value range of Kendall correlation coefficient τ is [-1,1], where a positive value of τ indicates that the change law of the operating parameter and the amplitude is positively correlated, and a negative value of τ indicates that the change law of the operating parameter and the amplitude is negatively correlated. The larger the absolute value of τ, the stronger the correlation between the operating parameter and the amplitude, and the smaller the absolute value, the weaker the correlation between the operating parameter and the amplitude. τ is 0, indicating that there is no correlation between the two parameters.

[0049] In one example, the operating parameters include at least two types, and step S101 includes: selecting an operating parameter whose absolute value of the Kendall correlation coefficient is greater than a second threshold as a target parameter.

[0050] Optionally, target parameters whose correlation coefficient with the stator winding end amplitude is greater than a set threshold are selected from the operating parameters as training data. This allows filtering of the operating parameters in the historical data, removing interference data with low correlation, and improving the quality of the training data, which is beneficial to improving the prediction accuracy of the subsequent stator winding end vibration fault diagnosis model.

[0051] Step S102, iteratively training a radial basis function generalized regression model based on training data and a gradient descent method, obtaining a theoretical value of the amplitude of the stator winding end, and calculating an error between the theoretical value and the true value.

[0052] Optionally, the radial basis function generalized regression model is trained and evaluated according to the training data obtained in step S101, and a training set and a validation set are divided according to a set ratio. The specific ratio is set according to the actual training needs. In this application, 80% of the training data is input into the radial basis function generalized regression model for training, and 20% of the training data is used to evaluate the model prediction effect. In this application, the gradient descent method is used as the training algorithm, and the learning rate and regularization parameters are dynamically adjusted during the training process, and the training error and convergence are monitored in real time.

[0053] In one example, the radial basis function generalized regression model uses a Gaussian permutation function as a kernel function.

[0054] Optionally, a Gaussian permutation function is used as the kernel function of the radial basis function generalized regression model, and the number, center, width and weight initialization method of the radial basis function are determined according to the characteristics and scale of the training data. Among them, the initialization number of radial basis functions needs to be determined according to the specific problem and data characteristics, usually set by a priori values ​​or by a clustering algorithm to determine the center point of the radial basis function in the input data, thereby indirectly determining the number of functions. The initial center point of the radial basis function can be determined by random selection, clustering algorithm, and self-organizing mapping. The width of the radial basis function controls the coverage of the function, affecting the response speed of the function and the corresponding nonlinear characteristics. The initial width is generally set manually according to the priori value or using a data-based method, and is dynamically set according to the variance of the data or the distance information in the feature space. The initial weight can be initialized by random initialization and linear regression initialization.

[0055] Preferably, in one example, the radial basis function generalized regression model clusters the features of the training data through self-organizing mapping to determine the number of sets, the center point and radius of each set.

[0056] Optionally, a self-organizing map is used to cluster the input data and map similar data features to adjacent neurons, thereby determining the center point of the radial basis function in an unsupervised manner. The traditional radial basis function network requires the number of hidden layer neurons to be manually set when it is constructed, which affects the performance and complexity of the network. After the introduction of self-organizing maps, the radial basis function center can be dynamically determined based on the natural structure of the data. Each set corresponds to a radial basis function center point, and the center point and radius of the set are determined at the same time. Dynamic adjustment can better adapt to the complexity of the data and avoid overfitting or underfitting problems.

[0057] In one example, the radial basis function generalized regression model includes an input layer, a hidden layer, and an output layer. The hidden layer uses a Gaussian permutation function as a kernel function for spatial mapping. The Gaussian permutation function is expressed as:

[0058]

[0059] Among them, x i represents the i-th data in the training data, c j represents the center of the jth set, δ j represents the diffusion coefficient of the jth set.

[0060] Step S103, in response to the error being less than the first threshold or the number of iterations being equal to the preset number of iterations, the iteration is stopped to obtain a stator winding end vibration fault diagnosis model.

[0061] Optionally, a gradient descent method is used as a training algorithm, and the learning rate and regularization parameters are dynamically adjusted during the training process. The training error and convergence are monitored in real time, and the training is stopped when the training error reaches a first threshold or the number of iterations reaches a preset number of iterations.

[0062] In summary, the present application establishes a dynamic stator winding end vibration fault diagnosis model through a radial basis function generalized regression model. Through the Kendall correlation coefficient, the parameters with high correlation with the stator winding end vibration are determined from the operating parameters, and the redundant data are filtered out, which reduces the complexity of the input information and the model and improves the working efficiency of the model. Compared with the traditional monitoring method that relies on sensors to collect vibration data and perform fixed threshold alarms, the diagnostic model provided by the present application can more accurately predict the theoretical amplitude of the stator winding end under different working conditions and realize early fault warning. It helps to intervene and repair the problem before it expands, reduce the risk of unexpected downtime, reduce maintenance costs and downtime, optimize the process and resource allocation of operation and maintenance management, and improve the accuracy and reliability of fault detection and the safety and stability of motor operation.

[0063] In a second aspect, an embodiment of the present application provides a method for diagnosing a stator winding end vibration fault. Figure 2 is a flow chart of a stator winding end vibration fault diagnosis method provided according to an embodiment of the present application, such as Figure 2 As shown, the method includes:

[0064] Step S201, obtaining real-time operating parameters of the steam turbine generator, and obtaining a theoretical value of the amplitude of the stator winding end according to the real-time operating parameters through the stator winding end vibration fault diagnosis model of the first aspect.

[0065] Step S202, calculating the deviation between the actual value of the amplitude monitored by the sensor and the theoretical value, and judging whether there is any abnormality in the vibration of the stator winding end according to the deviation.

[0066] In one example, step S202 includes: calculating a ratio of the deviation to the theoretical value, and in response to the ratio being greater than a third threshold, determining that an abnormality exists.

[0067] Optionally, it is possible to determine whether there is an abnormality based on the ratio of the deviation to the theoretical value. For example, when the ratio of the deviation to the theoretical value exceeds 10%, it is considered that there is an abnormality in the vibration of the stator winding end. At the same time, the abnormality level can also be determined based on the ratio of the deviation to the theoretical value. For example, when the ratio of the deviation to the theoretical value is 10% to 20%, it is a first-level abnormality, when the ratio of the deviation to the theoretical value is 20% to 40%, it is a second-level abnormality, and when the ratio of the deviation to the theoretical value is 40% to 50%, it is a third-level abnormality. The higher the abnormality level, the greater the corresponding risk. The third threshold value and the division of the abnormality level can be set according to the data characteristics during actual application, and are not limited to the data and methods listed in this application. In addition, emergency abnormalities can be given priority feedback to relevant professionals based on the risk level, so as to facilitate timely handling of abnormalities with higher risk levels to avoid causing greater losses.

[0068] In summary, the present application establishes a dynamic stator winding end vibration fault diagnosis model through a radial basis function generalized regression model, replacing the traditional monitoring method that relies on sensors to collect vibration data and perform fixed threshold alarms. The diagnostic model provided by the present application can more accurately predict the theoretical amplitude of the stator winding end under different working conditions and achieve early fault warning. It helps to intervene and repair problems before they expand, reduce the risk of unexpected downtime, reduce maintenance costs and downtime, optimize the process and resource allocation of operation and maintenance management, and improve the accuracy and reliability of fault detection and the safety and stability of motor operation.

[0069] In a third aspect, an embodiment of the present application provides a stator winding end vibration fault diagnosis model acquisition system, Figure 3 is a structural block diagram of a stator winding end vibration fault diagnosis model acquisition system provided in an embodiment of the present application, such as Figure 3 As shown, the system includes:

[0070] Acquisition module 100: used to acquire historical operation data, the historical operation data includes operation parameters and stator winding end amplitude at corresponding moments, and determine target parameters from the operation parameters through Kendall correlation coefficient, and determine training data based on the target parameters and the corresponding stator winding end amplitude.

[0071] Training module 200: used to iteratively train the radial basis function generalized regression model based on training data and gradient descent method, obtain the theoretical value of the amplitude of the stator winding end, and calculate the error between the theoretical value and the true value.

[0072] Model module 300: used to stop iteration in response to the error being less than a first threshold or the number of iterations being equal to a preset number of iterations, so as to obtain a stator winding end vibration fault diagnosis model.

[0073] In one example, the radial basis function generalized regression model in the training module 200 is configured to use a Gaussian permutation function as a kernel function.

[0074] In one example, the radial basis function generalized regression model in the training module 200 is configured to cluster the features of the training data through self-organizing mapping, and determine the number of sets, the center point and radius of each set.

[0075] In one example, the radial basis function generalized regression model in the training module 200 includes an input layer, a hidden layer, and an output layer. The hidden layer is configured to perform spatial mapping using a Gaussian permutation function as a kernel function. The Gaussian permutation function is expressed as:

[0076]

[0077] Among them, x i represents the i-th data in the training data, c j represents the center of the jth set, δ j represents the diffusion coefficient of the jth set.

[0078] In one example, the Kendall correlation coefficient in the acquisition module 100 represents the correlation between the operating parameter and the amplitude, and the calculation formula of the Kendall correlation coefficient is as follows:

[0079]

[0080] Among them, τ represents the Kendall correlation coefficient, N is the number of historical operation data, C represents the number of element logarithms of each operation parameter and the change law of amplitude, and D represents the number of element logarithms of each operation parameter and the change law of amplitude.

[0081] In one example, the operating parameters in the acquisition module 100 include at least two types, and the acquisition module 100 includes: selecting an operating parameter whose absolute value of the Kendall correlation coefficient is greater than a second threshold as a target parameter.

[0082] In summary, the present application establishes a dynamic stator winding end vibration fault diagnosis model through a radial basis function generalized regression model. Through the Kendall correlation coefficient, the parameters with high correlation with the stator winding end vibration are determined from the operating parameters, and the redundant data are filtered out, which reduces the complexity of the input information and the model and improves the working efficiency of the model. Compared with the traditional monitoring method that relies on sensors to collect vibration data and perform fixed threshold alarms, the diagnostic model provided by the present application can more accurately predict the theoretical amplitude of the stator winding end under different working conditions and realize early fault warning. It helps to intervene and repair the problem before it expands, reduce the risk of unexpected downtime, reduce maintenance costs and downtime, optimize the process and resource allocation of operation and maintenance management, and improve the accuracy and reliability of fault detection and the safety and stability of motor operation.

[0083] In a fourth aspect, an embodiment of the present application provides an electronic device, Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the method for obtaining a stator winding end vibration fault diagnosis model provided in the first aspect is implemented. Figure 4 The electronic device 60 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0084] The electronic device 60 may be in the form of a general-purpose computing device, for example, it may be a server device. The components of the electronic device 60 may include but are not limited to: at least one processor 61, at least one memory 62, and a bus 63 connecting different system components (including the memory 62 and the processor 61).

[0085] The bus 63 includes a data bus, an address bus, and a control bus.

[0086] The memory 62 may include a volatile memory, such as a random access memory (RAM) 621 and / or a cache memory 622 , and may further include a read-only memory (ROM) 623 .

[0087] The memory 62 may also include a program / utility 625 having a set (at least one) of program modules 624, such program modules 624 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0088] The processor 61 executes various functional applications and data processing by running the computer program stored in the memory 62, such as the method for acquiring the stator winding end vibration fault diagnosis model provided in the first aspect of the present application.

[0089] The electronic device 60 may also communicate with one or more external devices 64 (e.g., keyboards, pointing devices, etc.). Such communication may be performed via an input / output (I / O) interface 65. Furthermore, the model-generated device 60 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 66. As shown, the network adapter 66 communicates with other modules of the model-generated device 60 via a bus 63. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the model-generated device 60, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (RAID) systems, tape drives, and data backup storage systems.

[0090] It should be noted that although several units / modules or sub-units / modules of the electronic device are mentioned in the above detailed description, this division is merely exemplary and not mandatory. In fact, according to an embodiment of the present invention, the features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided into multiple units / modules to be embodied.

[0091] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described 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.

[0092] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for obtaining a stator winding end vibration fault diagnosis model, characterized in that: include: Acquire historical operation data, the historical operation data including operation parameters and stator winding end amplitudes at corresponding moments, determine target parameters from the operation parameters by using Kendall correlation coefficients, and determine training data according to the target parameters and the corresponding stator winding end amplitudes; Iteratively training a radial basis function generalized regression model based on the training data and the gradient descent method to obtain a theoretical value of the stator winding end amplitude, and calculating an error between the theoretical value and the true value; In response to the error being less than a first threshold or the number of iterations being equal to a preset number of iterations, the iteration is stopped to obtain a stator winding end vibration fault diagnosis model.

2. The method for obtaining a stator winding end vibration fault diagnosis model according to claim 1, characterized in that: The radial basis function generalized regression model is configured to use a Gaussian permutation function as a kernel function.

3. The method for obtaining a stator winding end vibration fault diagnosis model according to claim 2, characterized in that: The radial basis function generalized regression model is configured to cluster the features of the training data through self-organizing mapping, and determine the number of sets, the center point and radius of each set.

4. The method for obtaining a stator winding end vibration fault diagnosis model according to claim 3, characterized in that: The radial basis function generalized regression model includes an input layer, a hidden layer and an output layer, wherein the hidden layer is configured to perform spatial mapping using a Gaussian permutation function as a kernel function, and the Gaussian permutation function is expressed as: Among them, x i represents the i-th data in the training data, c j represents the center of the jth set, δ j represents the diffusion coefficient of the jth set.

5. The method for obtaining a stator winding end vibration fault diagnosis model according to claim 1, characterized in that: The Kendall correlation coefficient represents the correlation between the operating parameter and the amplitude. The calculation formula of the Kendall correlation coefficient is as follows: Wherein, τ represents the Kendall correlation coefficient, N is the number of the historical operation data, C represents the number of element logarithms of each operation parameter being positively correlated with the change pattern of the amplitude, and D represents the number of element logarithms of each operation parameter being negatively correlated with the change pattern of the amplitude.

6. The method for obtaining a stator winding end vibration fault diagnosis model according to claim 5, characterized in that: The operating parameters include at least two types, and determining the target parameters from the operating parameters includes: The operating parameter whose absolute value of the Kendall correlation coefficient is greater than the second threshold is selected as the target parameter.

7. A method for diagnosing stator winding end vibration faults, characterized in that: include: Acquire the real-time operating parameters of the steam turbine generator, and obtain the theoretical value of the stator winding end amplitude according to the real-time operating parameters through the stator winding end vibration fault diagnosis model described in any one of claims 1 to 6; The deviation between the actual value of the amplitude monitored by the sensor and the theoretical value is calculated, and whether there is abnormality in the vibration of the stator winding end is determined based on the deviation.

8. The stator winding end vibration fault diagnosis method according to claim 7, characterized in that: The step of judging whether the vibration of the stator winding end is abnormal according to the deviation includes: A ratio of the deviation to the theoretical value is calculated, and in response to the ratio being greater than a third threshold, it is determined that an abnormality exists.

9. A stator winding end vibration fault diagnosis model acquisition system, characterized in that: include: Acquisition module: used to acquire historical operation data, the historical operation data includes operation parameters and stator winding end amplitude at corresponding moments, determine target parameters from the operation parameters through Kendall correlation coefficient, and determine training data according to the target parameters and the corresponding stator winding end amplitude; Training module: used for iteratively training the radial basis function generalized regression model based on the training data and the gradient descent method, obtaining the theoretical value of the amplitude of the stator winding end, and calculating the error between the theoretical value and the true value; Model module: used to stop iteration in response to the error being less than a first threshold or the number of iterations being equal to a preset number of iterations, so as to obtain a stator winding end vibration fault diagnosis model.

10. An electronic device, characterized in that: include Memory, processor, and A computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for obtaining a stator winding end vibration fault diagnosis model as described in any one of claims 1 to 6 is implemented.

Images