Steam turbine set vibration fault diagnosis method and system based on naive Bayes classification model
Through the improved naive Bayes classification model, combining symmetric uncertainty and approximate Markov blanket method, and using interpolation difference method, the problem of insufficient accuracy of steam turbine fault diagnosis in the prior art is solved, and higher classification accuracy is achieved rate and fault diagnosis accuracy.
Patent Information
- Application Number
- CN202411232251.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-04
- Publication Date
- 2025-01-07
AI Technical Summary
The existing turbine fault diagnosis methods, especially those based on support vector machines, are difficult to achieve ideal prediction classification accuracy in parameter optimization, and the particle swarm algorithm is prone to fall into local optimality and has low search accuracy.
Using the improved naive Bayes classification model, the correlation between fault attributes and categories is calculated by symmetric uncertainty methods, the final weight of each fault attribute is determined using the approximate Markov blanket method, and it is introduced into the interpolation difference method. A property-weighted interpolated differential distance metric method (AWIVDM) is formed to improve classification accuracy.
It improves the accuracy of distance measurement and classification accuracy, effectively removes redundant fault attributes, and improves the accuracy of vibration fault diagnosis of steam turbine units.
Smart Images
Figure CN119271994A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of pattern recognition and machine intelligence, and in particular to a method and system for diagnosing vibration faults of a steam turbine unit based on a naive Bayesian classification model. Background Art
[0002] Steam turbine is one of the most important equipment in power generation enterprises. Whether it can operate safely and stably affects the safety and economy of the entire generator set. For steam turbine, vibration failure is the most important equipment failure. Being able to handle steam turbine vibration failure in a timely and reasonable manner is an important guarantee for the safe and stable production of power plants.
[0003] In recent years, many artificial intelligence methods have been applied to the diagnosis of steam turbine faults, including support vector machines, expert systems, particle swarm algorithms, neural networks, etc. They have all achieved some results. Support vector machines can be used to diagnose steam turbine faults. Support vector machines require less training sample data, have strong generalization ability, and are widely used, but the regularization coefficient is difficult to determine. If the parameters cannot be optimized, it is difficult to obtain the ideal prediction and classification accuracy. It has also been proposed to use particle swarm algorithms to optimize the parameters of support vector machines, but particle swarm algorithms are prone to fall into local optimal points and the search accuracy is not high. Summary of the invention
[0004] The purpose of the present invention is to overcome the deficiencies in the prior art and to provide a method and system for diagnosing vibration faults of a steam turbine unit based on a naive Bayesian classification model. By using an improved naive Bayesian classification model, the accuracy of distance measurement can be improved, thereby further improving the classification accuracy and achieving the accuracy of steam turbine unit vibration fault diagnosis.
[0005] To achieve the above object, the present invention is implemented by adopting the following technical solutions:
[0006] On the one hand, the present invention provides a method for diagnosing vibration faults of a steam turbine unit based on a naive Bayesian classification model, comprising:
[0007] Obtain vibration data of steam turbine unit;
[0008] According to the vibration data of the steam turbine unit, based on a pre-trained naive Bayes classification model, a vibration fault diagnosis result of the steam turbine unit is obtained;
[0009] The training of the naive Bayes classification model includes:
[0010] The training samples of historical steam turbine unit vibration data are input into the naive Bayes classification model, and the correlation between each fault attribute and the fault category on the training sample is calculated using the symmetric uncertainty method to obtain the initial weight of each fault attribute. The approximate Markov blanket method is used to determine the final weight of each fault attribute; the final weight is introduced into the inserted value difference method to obtain the trained naive Bayes classification model.
[0011] Optionally, the symmetric uncertainty method is used to calculate the correlation of each fault attribute category on the training sample, such as formula (1) and (2):
[0012] SU x l c = 2 [ IG ( x l | c ) H ( x l ) + H ( x l | c ) ] (1);
[0013] (2);
[0014] in, The correlation between each fault attribute and the fault category; For the Fault attributes; is the fault category of the training sample; is the information gain; is information entropy; is the conditional entropy.
[0015] Optionally, the final weight of each fault attribute is determined using an approximate Markov blanket method, such as formulas (3) and (4):
[0016] (3);
[0017] (4);
[0018] in, For the The initial weights of the fault attributes; For the The final weight of each fault attribute; For the Fault attributes; is the number of fault attributes in the training samples; is the fault category of the training sample; The fault attribute correlation value is greater than The set of fault attributes, is the preset threshold; is the correlation between each fault attribute and the fault category in the training sample.
[0019] Optionally, the insertion value difference method of the final weight is introduced, such as formula (5) and (6):
[0020] (5);
[0021] (6);
[0022] For the The interpolation difference distance metric of training samples; To use The distance between the training samples and the test samples; For the Fault attributes; is the number of fault attributes in the training samples; For the The first of the training samples The distance between the fault attribute and the test sample; is the first value obtained using the interpolation difference method. The first of the training samples The distance between the fault attribute and the test sample; , are the values of the fault attributes in the training samples and the test samples respectively; For the The final weight of each fault attribute; is the fault category of the training sample; is the total number of fault categories of training samples; is the number of fault categories; For the The value of the fault category to which the training samples belong; , They are , Belong to The probability of a class.
[0023] Optionally, the trained naive Bayes classification model is as follows:
[0024] (7);
[0025] (8);
[0026] in, is the trained naive Bayes classification model; For the The weight of the training samples; is a linear weight function; For the The interpolation difference distance metric of training samples, For the The distance metric of the interpolation difference of training samples satisfies ; To satisfy The number of training samples; is the total number of training samples.
[0027] Optionally, setting a test sample of the historical steam turbine unit vibration data, and determining a correct classification category of the test sample according to the trained naive Bayes classification model;
[0028] Calculate the posterior probability of the training sample category, take the training sample category with the largest posterior probability as the correct category for the test sample, and take the fault attribute value as The posterior probability of the training sample category is calculated as shown in formulas (9)-(11):
[0029] (9);
[0030] (10);
[0031] (11);
[0032] in, is the posterior probability of the training sample category; is the prior probability of the training sample category; For value The conditional probability of is the number of fault categories; is the total number of training sample categories; is the total number of training samples; is the trained naive Bayes classification model; For the The value of the fault category to which the training samples belong; For the The value of the fault category to which the training samples belong; is the fault category of the training sample; is the indicator function; Fault Attributes The total number of possible values; , are the values of the fault attributes in the training samples and the test samples respectively.
[0033] Optionally, calculating the classification accuracy of the historical steam turbine unit vibration data according to the correct classification of the test sample includes:
[0034] If the number of test sample categories of the historical steam turbine unit vibration data is n, and the test samples are classified and tested according to the trained naive Bayes classification model, and the number of correctly classified categories is m, then the classification accuracy of the historical steam turbine unit vibration data is m / n*100%.
[0035] In a second aspect, the present invention provides a steam turbine unit vibration fault diagnosis system based on a naive Bayes classification model, comprising:
[0036] A data acquisition module, used for acquiring vibration data of the steam turbine unit;
[0037] Model training module, used to train the naive Bayes classification model;
[0038] The classification calculation module is used to calculate the classification accuracy of historical steam turbine unit vibration data.
[0039] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the steam turbine unit vibration fault diagnosis method based on the naive Bayes classification model described in the first aspect.
[0040] In a fourth aspect, the present invention provides a computer device, comprising:
[0041] Memory, for storing computer instructions;
[0042] A processor is used to execute the computer instructions to implement the steps of the steam turbine unit vibration fault diagnosis method based on the naive Bayes classification model described in the first aspect.
[0043] Beneficial Effects
[0044] Compared with the prior art, the present invention has the following beneficial effects:
[0045] The present invention can improve the accuracy of distance measurement through an improved naive Bayesian classification model, thereby further improving the classification accuracy and achieving the accuracy of turbine unit vibration fault diagnosis; the naive Bayesian classification model of the present invention determines the role weights of various fault attributes of the fault in sample distance calculation by introducing an approximate Markov blanket method, and can accurately calculate the weights of training samples; in the approximate Markov blanket method, the calculation of the correlation between fault attributes is obtained by a symmetric uncertainty method, and the use of the symmetric uncertainty method can effectively measure the correlation size of various fault attributes of the fault, and at the same time remove a part of redundant fault attributes, thereby effectively determining the influence weight of each fault attribute value in calculating the sample distance; the improved fault attribute weight is combined with an interpolation difference method (IVDM) to obtain an interpolation difference distance measurement method (AWIVDM) based on attribute weighting, and the naive Bayesian classification model of the present invention is obtained according to the AWIVDM method. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] Figure 1 FIG. 1 is a flow chart of a steam turbine unit vibration fault diagnosis method based on a naive Bayesian classification model in an embodiment of the present invention;
[0047] Figure 2 FIG. 1 is a schematic structural diagram of a steam turbine unit vibration fault diagnosis system based on a naive Bayesian classification model in an embodiment of the present invention;
[0048] Figure 3 Shown is a schematic diagram of steam turbine unit vibration fault diagnosis in an embodiment of an existing naive Bayes classification model;
[0049] Figure 4 Shown is a schematic diagram of steam turbine unit vibration fault diagnosis in one embodiment of the naive Bayesian classification model of the present invention;
[0050] Figure 5 Shown is a schematic diagram of vibration data of a steam turbine unit in one embodiment of the present invention. DETAILED DESCRIPTION
[0051] The technical solution of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations on the technical solution of the present invention. The embodiments of the present invention and the technical features in the embodiments may be combined with each other unless there is a conflict.
[0052] The term "and / or" is only a description of the association relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " generally indicates that the related objects are in an "or" relationship.
[0053] Example 1
[0054] like Figure 1 As shown, this embodiment introduces a method for diagnosing vibration faults of a steam turbine unit based on a naive Bayesian classification model, and the specific steps include:
[0055] Obtain the vibration data of the steam turbine unit. Figure 5 The vibration data obtained in this embodiment is shown.
[0056] According to the vibration data of the steam turbine unit, based on a pre-trained naive Bayesian classification model, a vibration fault diagnosis result of the steam turbine unit is obtained.
[0057] The training of the naive Bayes classification model includes:
[0058] The training samples of the acquired historical steam turbine unit vibration data are input into the Naive Bayes classification model.
[0059] Initialize threshold settings In this embodiment, .
[0060] The symmetric uncertainty method is used to calculate the correlation between each fault attribute and the fault category on the training sample, the value of the correlation is used as the initial weight of each fault attribute, and the approximate Markov blanket method is used to determine the final weight of each fault attribute.
[0061] The symmetric uncertainty method is a method based on nonlinear correlation measurement, which can effectively measure the strength of the mutual dependence of any two random variables. The symmetric uncertainty is defined as follows:
[0062] SU XY = 2 [ IG ( X | Y ) H ( X ) + H ( X | Y ) ] ;
[0063] ;
[0064] in, , are two arbitrary variables; is the information gain; is information entropy; is the conditional entropy; For a given After the variable, Reduction in variable certainty. The larger the variable and The stronger the correlation is. The determination of fault attribute weights first measures the correlation between fault attributes of the same category. According to the value of class correlation, the weights of irrelevant fault attributes and fault attributes with strong dependencies are set to 0. Irrelevant fault attributes are determined by setting thresholds. To determine, the strongly dependent fault attributes are determined using the approximate Markov blanket method, and the approximate Markov blanket is defined as follows:
[0065] Approximate Markov Blanket: Any Fault Attribute Class-related applications measure, Same fault attributes Redundancy Metrics, if yes The Markov blanket satisfies:
[0066] and .
[0067] In this embodiment, the symmetric uncertainty method is used to calculate the correlation between each fault attribute and the fault category on the training sample, such as formula:
[0068] SU x l c = 2 [ IG ( x l | c ) H ( x l ) + H ( x l | c ) ] ;
[0069] ;
[0070] in, The correlation between each fault attribute and the fault category; For the Fault attributes; is the fault category of the training sample; is the information gain; is information entropy; is the conditional entropy.
[0071] The approximate Markov blanket method is used to add weights to the initial weights of each fault attribute, and the weights after adding weights are used as the final weights of each fault attribute. The weight of each fault attribute is as follows:
[0072] ;
[0073] ;
[0074] in, For the The initial weights of the fault attributes; For the The final weight of each fault attribute; For the Fault attributes; is the number of fault attributes in the training samples; is the fault category of the training sample; is the correlation between each fault attribute and the fault category on the training sample; The fault attribute correlation value is greater than The set of fault attributes, is the preset threshold, The value range is In actual situations, if the data set contains many irrelevant fault attributes, Try to get it smaller. If the value is too large, some relevant fault attributes may be lost. If there are many relevant fault attributes, It can be obtained larger, and some strongly dependent fault attributes can be removed first.
[0075] The final weight is introduced into the insertion value difference method to obtain a trained naive Bayes classification model.
[0076] The role of each fault attribute in calculating the weight of the training sample is fully considered, and the insertion value difference method of the final weight is introduced, such as the formula:
[0077] ;
[0078] ;
[0079] in, For the The interpolation difference distance metric of training samples; To use The distance between the training samples and the test samples; For the Fault attributes; is the number of fault attributes in the training samples; For the The first of the training samples The distance between the fault attribute and the test sample; is the first value obtained using the interpolation difference method. The first of the training samples The distance between the fault attribute and the test sample; , are the values of the fault attributes in the training samples and the test samples respectively; For the The final weight of each fault attribute; is the fault category of the training sample; is the total number of fault categories of training samples; is the number of fault categories; For the The value of the fault category to which the training samples belong; , They are , Belong to The probability of a class.
[0080] The VDM calculation method when calculating the discrete fault attribute sample distance is as follows:
[0081] .
[0082] For continuous fault attributes, the fault attributes are discretized using the equal width method. The fault attributes are discretized into equal intervals. The width is calculated as follows:
[0083] ;
[0084] ;
[0085] Among them, and are not the probability values after discretization of the fault attribute values. The probability values are calculated as follows:
[0086]
[0087] ;
[0088] in, and They are the midpoints of two continuous intervals after the continuous fault attributes are discretized.
[0089] The trained naive Bayes classification model is as follows:
[0090] ;
[0091] ;
[0092] in, is the trained naive Bayes classification model; For the The weight of the training samples; is a linear weight function; For the The interpolation difference distance metric of training samples, For the The distance metric of the interpolation difference of training samples satisfies ; To satisfy The number of training samples; is the total number of training samples. The sum of all sample weights is approximately equal to , The smaller the value of is, the farther away from the test sample is, and vice versa.
[0093] Setting a test sample of the historical steam turbine unit vibration data, and determining a correct classification category of the test sample according to the trained naive Bayes classification model;
[0094] Calculate the posterior probability of the training sample category, take the training sample category with the largest posterior probability as the correct category for the test sample, and take the fault attribute value as The posterior probability of the training sample category is calculated as shown in formulas (9)-(11):
[0095] ;
[0096] ;
[0097] ;
[0098] in, is the posterior probability of the training sample category; is the prior probability of the training sample category; For value The conditional probability of is the number of fault categories; is the total number of training sample categories; is the total number of training samples; is the trained naive Bayes classification model; For the The value of the fault category to which the training samples belong; For the The value of the fault category to which the training samples belong; is the fault category of the training sample; is the indicator function; Fault Attributes The total number of all possible values; , are the values of the fault attributes in the training samples and the test samples respectively.
[0099] According to the correct classification of the test sample, calculating the classification accuracy of the historical steam turbine unit vibration data includes:
[0100] If the number of test sample categories of the historical steam turbine unit vibration data is n, and the test samples are classified and tested according to the trained naive Bayes classification model, and the number of correctly classified categories is m, then the classification accuracy of the historical steam turbine unit vibration data is m / n*100%.
[0101] Example 2
[0102] like Figure 2 As shown, this embodiment introduces a steam turbine unit vibration fault diagnosis system based on a naive Bayes classification model, including:
[0103] A data acquisition module, used for acquiring vibration data of the steam turbine unit;
[0104] Model training module, used to train the naive Bayes classification model;
[0105] The classification calculation module is used to calculate the classification accuracy of historical steam turbine unit vibration data.
[0106] Example 3
[0107] like Figure 3 , 4 As shown, this embodiment introduces a specific experimental design of a steam turbine unit vibration fault diagnosis method based on a naive Bayes classification model:
[0108] Collect the vibration conditions of the steam turbine unit, extract the fault model data of the vibration of the generator bearings 1 to 8, extract the vibration feature data including rotor imbalance, dynamic and static friction, oil film vortex, rotor misalignment and normal operation, and create training and test samples. The specific steps are as follows:
[0109] Step 1: Determine the characteristic vector of vibration cause data;
[0110] Step 2: Select essential features from the feature vector that can distinguish the cause of vibration;
[0111] Step 3: Divide the selected essential feature vectors into two parts: training samples and test samples.
[0112] from Figure 3 and Figure 4 It can be seen that the classification accuracy of the improved naive Bayesian classification model (AWIVDM-LWNBC) on the data set is higher than that of the unimproved naive Bayesian classification model (IVDM-LWNBC). This is because in the process of determining the weight of the training samples, some redundant and irrelevant attributes are removed from the naive Bayesian classification model through attribute weighting, and the different effects of each attribute in distance measurement are considered. This shows that the attribute weighting method is effective in improving the local weighted naive Bayesian classification model.
[0113] The local weighted naive Bayes classification model adopts a lazy learning strategy, which selects some training samples to construct the naive Bayes classification model by setting the training sample weights, effectively relaxing the class conditional independence assumption. The distance measurement method based on probability measurement (IVDM) can effectively calculate data sets containing nominal and numerical attributes, but these two methods ignore the different roles of different attributes in distance measurement during the calculation process, as well as the influence of possible redundant and irrelevant attributes on distance measurement.
[0114] This paper proposes a method for vibration fault diagnosis of steam turbine units based on the naive Bayes classification model. Through an improved probabilistic distance measurement method (AWIVDM), the idea of Markov blanket and the correlation measurement method of symmetric uncertainty are used to add weights to each attribute to improve the accuracy of distance measurement, thereby further improving the classification accuracy and achieving the accuracy of vibration fault diagnosis of steam turbine units.
[0115] Example 4
[0116] This embodiment introduces a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the steam turbine unit vibration fault diagnosis method based on the naive Bayes classification model described in Example 1 are implemented.
[0117] Example 5
[0118] This embodiment introduces a computer device, including:
[0119] Memory, for storing computer instructions;
[0120] A processor is used to execute the computer instructions to implement the steps of the steam turbine unit vibration fault diagnosis method based on the naive Bayes classification model described in Example 1.
[0121] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.
[0122] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0123] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0124] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0125] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which all fall within the protection of the present invention.
Claims
1. A steam turbine unit vibration fault diagnosis method based on a naive Bayesian classification model, characterized in that: include: Obtain vibration data of steam turbine unit; According to the vibration data of the steam turbine unit, based on a pre-trained naive Bayes classification model, a vibration fault diagnosis result of the steam turbine unit is obtained; The training of the naive Bayes classification model includes: The training samples of historical steam turbine unit vibration data are input into the naive Bayes classification model, and the correlation between each fault attribute and the fault category on the training sample is calculated using the symmetric uncertainty method to obtain the initial weight of each fault attribute. The approximate Markov blanket method is used to determine the final weight of each fault attribute; the final weight is introduced into the inserted value difference method to obtain the trained naive Bayes classification model.
2. The steam turbine unit vibration fault diagnosis method based on the naive Bayes classification model according to claim 1 is characterized in that: The symmetric uncertainty method is used to calculate the correlation between each fault attribute and the fault category on the training sample, as shown in formulas (1) and (2): （1）； （2）； in, The correlation between each fault attribute and the fault category; For the Fault attributes; is the fault category of the training sample; is the information gain; is information entropy; is the conditional entropy.
3. The steam turbine unit vibration fault diagnosis method based on the naive Bayes classification model according to claim 1 is characterized in that: The approximate Markov blanket method is used to determine the final weight of each fault attribute, such as formula (3) and (4): （3）； （4）； in, For the The initial weights of the fault attributes; For the The final weight of each fault attribute; For the Fault attributes; is the number of fault attributes in the training samples; is the fault category of the training sample; The fault attribute correlation value is greater than The set of fault attributes, is the preset threshold; is the correlation between each fault attribute and the fault category in the training sample.
4. The steam turbine unit vibration fault diagnosis method based on the naive Bayes classification model according to claim 1 is characterized in that: The insertion value difference method of the final weight is introduced, such as formula (5) and (6): （5）； （6）； in, For the The interpolation difference distance metric of training samples; To use The distance between the training samples and the test samples; For the Fault attributes; is the number of fault attributes in the training samples; For the The first of the training samples The distance between the fault attribute and the test sample; is the first value obtained using the interpolation difference method. The first of the training samples The distance between the fault attribute and the test sample; , are the values of the fault attributes in the training samples and the test samples respectively; For the The final weight of each fault attribute; is the fault category of the training sample; is the total number of fault categories of training samples; is the number of fault categories; For the The value of the fault category to which the training samples belong; , They are , Belong to The probability of a class.
5. The steam turbine unit vibration fault diagnosis method based on the naive Bayes classification model according to claim 1 is characterized in that: The trained naive Bayes classification model is as shown in formulas (7) and (8): （7）； （8）； in, is the trained naive Bayes classification model; For the The weight of the training samples; is a linear weight function; For the The interpolation difference distance metric of training samples, For the The distance metric of the interpolation difference of training samples satisfies ; To satisfy The number of training samples; is the total number of training samples.
6. The steam turbine unit vibration fault diagnosis method based on the naive Bayes classification model according to claim 1 is characterized in that: Setting a test sample of the historical steam turbine unit vibration data, and determining a correct classification category of the test sample according to the trained naive Bayes classification model; Calculate the posterior probability of the training sample category, take the training sample category with the largest posterior probability as the correct category for the test sample, and take the fault attribute value as The posterior probability of the training sample category is calculated as shown in formulas (9)-(11): （9）； （10）； （11）； in, is the posterior probability of the training sample category; is the prior probability of the training sample category; For value The conditional probability of is the number of fault categories; is the total number of training sample categories; is the total number of training samples; is the trained naive Bayes classification model; For the The value of the fault category to which the training samples belong; For the The value of the fault category to which the training samples belong; is the fault category of the training sample; is the indicator function; Fault Attributes The total number of possible values; , are the values of the fault attributes in the training samples and the test samples respectively.
7. The steam turbine unit vibration fault diagnosis method based on the naive Bayes classification model according to claim 6 is characterized in that: According to the correct classification of the test sample, calculating the classification accuracy of the historical steam turbine unit vibration data includes: If the number of test sample categories of the historical steam turbine unit vibration data is n, and the test samples are classified and tested according to the trained naive Bayes classification model, and the number of correctly classified categories is m, then the classification accuracy of the historical steam turbine unit vibration data is m / n*100%.
8. A steam turbine unit vibration fault diagnosis system based on a naive Bayesian classification model, characterized in that: include: A data acquisition module, used for acquiring vibration data of the steam turbine unit; Model training module, used to train the naive Bayes classification model; The classification calculation module is used to calculate the classification accuracy of historical steam turbine unit vibration data.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the steam turbine unit vibration fault diagnosis method based on the naive Bayes classification model described in any one of claims 1 to 7 are implemented.
10. A computer device, characterized in that: include: Memory, for storing computer instructions; A processor is used to execute the computer instructions to implement the steps of the steam turbine unit vibration fault diagnosis method based on the naive Bayes classification model described in any one of claims 1-7.