A wind power equipment multi-parameter monitoring method and system based on AI

By constructing a multi-dimensional data matrix and correlation analysis, calculating the weight value of wind power parameters, the problem of weakening data characteristics caused by relying on empirical values ​​in the existing technology is solved, and the accuracy of operating status monitoring of wind power equipment and fault warning capabilities are significantly improved.

CN118855647BActive Publication Date: 2024-12-03国投检测科技(山东)有限公司 +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411327926.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-24
Publication Date
2024-12-03
Estimated Expiration
2044-09-24

AI Technical Summary

Technical Problem

In the multi-parameter monitoring of wind power equipment, the data characteristics are weakened by relying on empirical value allocation, affecting the accuracy of the artificial intelligence model for the operational status detection of wind power equipment.

Method used

By building a multi-dimensional data matrix, calculate the correlation and variable point detection between wind power parameters, generate correlation impact and fault representation values, calculate the weight values ​​of each type of wind power parameters, perform data fusion and train neural networks, and improve monitoring accuracy sex.

Benefits of technology

It significantly improves the accuracy of operating status monitoring of wind power equipment, can more comprehensively analyze and understand the complex relationships and patterns of wind power parameters, and enhances the early warning ability for faults.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118855647B_ABST
    Figure CN118855647B_ABST
Patent Text Reader

Abstract

The present application relates to the field of data processing technology, and specifically to a method and system for multi-parameter monitoring of wind power equipment based on AI, the method comprising: obtaining a multidimensional data matrix and an average data matrix according to all types of wind power parameters at all times under various operating states of the wind power equipment; analyzing the average data matrix, and obtaining various types of associated wind power parameters according to the differences of different types of wind power parameters; performing change point detection on various types of wind power parameters, and obtaining an average correlation coefficient according to the numerical changes of different types of wind power parameters between adjacent change points of various types; constructing an associated influence degree; obtaining a fault representation value according to the differences of various types between different operating states; and fusing all wind power parameters at each time in each multidimensional data matrix. The present application aims to fuse different wind power parameters and improve the accuracy of neural network monitoring of the operating status of wind power equipment.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and specifically to an AI-based multi-parameter monitoring method and system for wind power equipment. Background Art

[0002] Wind power equipment is equipment that uses wind energy to generate electricity and is an important part of the wind power industry. In order to improve the operating reliability of wind turbines, reduce maintenance costs, and ensure the operating efficiency and safety of wind farms, wind power equipment needs to be monitored. With the rapid development of wind power technology, the demand for integrated and intelligent monitoring and management systems is growing, requiring the ability to accurately identify the status of equipment under various complex working conditions, and achieve early warning of faults and proactive maintenance.

[0003] With the development of artificial intelligence, artificial intelligence algorithms can be used to monitor various parameters of wind power equipment and identify abnormal patterns. The amount of wind power equipment data is large and diverse. Before training the model, it is usually necessary to use data fusion technology to fuse the data. The current fusion method usually relies on experience to assign weights to different types of data. This practice may weaken the characteristics of important data, thereby affecting the accuracy of the artificial intelligence model's detection results of the wind power equipment operating status. Summary of the invention

[0004] In view of the above, it is necessary to provide an AI-based multi-parameter monitoring method and system for wind power equipment to solve the above problems.

[0005] The first aspect of the present application provides a multi-parameter monitoring method for wind power equipment based on AI, the method comprising:

[0006] According to all types of wind power parameters at all times in a preset time period under various operating conditions of the wind power equipment, a multidimensional data matrix under various operating conditions is obtained; for various operating conditions, the average values ​​of all data at the same position of a preset number of multidimensional data matrices are used to form an average data matrix for various operating conditions;

[0007] Based on the differences between various types of wind power parameters and other types in each average data matrix, various types of associated type wind power parameters are obtained; change point detection is performed on various types of wind power parameters in each average data matrix, and the numerical changes of the associated type wind power parameters between adjacent change points of various types are analyzed to obtain the average association coefficients of various types in each average data matrix; the number of associated types and the average association coefficients of various types are integrated to obtain the association influence of various types;

[0008] According to the various types of differences between the average data matrices of different operating states, various types of fault representation values ​​are obtained; according to the various types of fault representation values ​​and the associated influence, all wind power parameters at each moment in each multidimensional data matrix are integrated;

[0009] A neural network is trained according to the fusion results of wind power parameters at each moment in each multidimensional data matrix and the corresponding operating status; and the operating status of the current wind power equipment is monitored according to the trained neural network.

[0010] The multidimensional data matrix obtained under various operating conditions is specifically:

[0011] Each row in the multidimensional data matrix represents the wind power parameters of one type at all times within a preset time period; each column represents the wind power parameters of all types at the same time.

[0012] The various types of associated wind power parameters are obtained as follows:

[0013] Based on the distance metric between any two types of wind power parameters in each average data matrix, determining whether there is a correlation between the any two types of wind power parameters;

[0014] All types of wind power parameters associated with various types are regarded as associated type wind power parameters of various types.

[0015] The specific process of determining whether any two types of wind power parameters are correlated is as follows:

[0016] The distance metric between any two types of wind power parameters in each average data matrix is ​​calculated and normalized. When the distance metric is greater than a preset correlation threshold, it is determined that there is correlation between the any two types of wind power parameters; otherwise, there is no correlation.

[0017] The method of obtaining the average correlation coefficients of various types in each average data matrix includes:

[0018] For each average data matrix, change point detection is performed on wind power parameters of various types at all times to obtain several trend change points;

[0019] The correlation coefficient of the kth trend change point of the i-th type is recorded as , its formula form is: ;in, Indicates the number of associated types of the i-th type; , They represent the k+1th trend change point and the kth trend change point of the i-th type respectively; , Respectively represent the rth association type of the i-th type , Wind power parameters at the corresponding time; Indicates a preset value greater than zero;

[0020] The average level of the correlation coefficients of all trend change points of various types is taken as the average correlation coefficient of various types.

[0021] The steps of obtaining various types of fault representation values ​​are as follows:

[0022] The distances of various types of wind power parameters between any two operating states are obtained, and the average levels of all pairwise combinations of the distances are taken as various types of fault representation values.

[0023] The fusion of all wind power parameters at each moment in each multidimensional data matrix is ​​specifically as follows:

[0024] The normalized value of the ratio of each type of fault representation value to the associated impact degree is used as the weight value of each type;

[0025] For each moment in each multidimensional data matrix, the product of each type of wind power parameter and its corresponding weight value is calculated, and the cumulative sum of the products of all types of wind power parameters is used as the fusion result of the wind power parameters at each moment in each multidimensional data matrix.

[0026] The training of the neural network is specifically as follows:

[0027] The fusion results of wind power parameters at each moment in the multidimensional data matrix under each operating state are calculated respectively, and the fusion results are marked according to the corresponding operating state. All marked fusion results are used as input of the neural network model to train the neural network model.

[0028] The process of monitoring the current operating status of the wind power equipment includes:

[0029] The result of fusing all wind power parameters at the current moment is used as the input of the trained neural network model to output the monitoring result of the current wind power equipment operating status.

[0030] In the second aspect, an embodiment of the present application also provides an AI-based multi-parameter monitoring system for wind power equipment, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above-mentioned methods when executing the computer program.

[0031] In the above scheme, by analyzing all types of wind power parameters at all times under various operating conditions of wind power equipment, a multidimensional data matrix is ​​established, which is helpful for comprehensive and in-depth analysis and understanding of the complex relationships and patterns of wind power parameters; further obtaining an average data matrix is ​​helpful for comprehensive analysis of various types of data under the same operating state; according to the differences between different types of wind power parameters in the average data matrix, various types of related wind power parameters are obtained, and based on the correlation characteristics between different types of wind power parameters, an average correlation coefficient is constructed, and by comprehensively analyzing the correlation between different types of data, it is evaluated whether these data are affected by mutual correlation, thereby affecting the accuracy of operating status monitoring; further constructing the correlation coefficient The influence degree helps to understand the relationship between various data types and determine how they jointly affect the accuracy of wind power equipment operating status monitoring; based on the various types of differences between the average data matrices of different operating states, a fault representation value is constructed, which has the beneficial effect of measuring the changing characteristics of various types of wind power parameters under different operating states, and is used to characterize the degree to which the data represents different operating states; based on the associated influence degree and fault representation value, the weight value of each type of wind power parameter is calculated, and by assigning different weights to each type of wind power parameter and fusing the data at each moment, it can more accurately reflect the operating status of the wind power equipment, and then use these fused data to train the neural network model, thereby significantly improving the accuracy of operating status monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 A flowchart of a method for multi-parameter monitoring of wind power equipment based on AI provided in one embodiment of the present application;

[0033] Figure 2 A schematic diagram of a CPD change point detection algorithm provided in one embodiment of the present application;

[0034] Figure 3 A flowchart for obtaining fusion results provided in one embodiment of the present application. DETAILED DESCRIPTION

[0035] In the description of the embodiments of the present application, words such as "exemplary", "or", "for example" and the like are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary", "or", "for example" and the like is intended to present related concepts in a concrete manner.

[0036] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art in the present application. The terms used in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application.

[0037] It should also be noted that the terms "first" and "second" in this application and the accompanying drawings are used to distinguish similar objects, rather than to describe a specific order or sequence. The method disclosed in the embodiments of the present application or the method shown in the flow chart includes one or more steps for implementing the method. Without departing from the scope of protection of the present application, the execution order of multiple steps can be interchanged with each other, and some steps can also be deleted.

[0038] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0039] The following is a detailed description of a specific solution of an AI-based wind power equipment multi-parameter monitoring method and system provided by the present application in conjunction with the accompanying drawings.

[0040] See also Figure 1 , which shows a flowchart of a method for multi-parameter monitoring of wind power equipment based on AI provided by an embodiment of the present application, the method comprising the following steps:

[0041] The first step: based on all types of wind power parameters at all times in a preset time period under various operating conditions of the wind power equipment, obtain a multidimensional data matrix under various operating conditions; for various operating conditions, the average value of all data at the same position of a preset number of multidimensional data matrices is used to form an average data matrix for various operating conditions.

[0042] This application collects wind power parameters in the operation of wind power equipment through a wind power generation environment monitoring system to monitor these faults. The types of wind power parameters collected in this application include: wind speed, generator speed, gearbox temperature, generator current, power generation, blade vibration displacement, blade vibration speed, blade vibration frequency, wind rotor bearing speed, etc.

[0043] Since the operating states of wind power equipment are divided into four types: normal operation, generator failure, blade failure, and both generator and blade failure. For each operating state, the wind power parameters of the preset time period are collected. It should be noted that the total duration of the preset time period is set to 30 minutes, and equal-interval sampling is performed within the preset time period, and the sampling interval is set to 5 seconds. In the preset time period, all types of wind power parameters at various operating states of the wind power equipment at each moment are sampled at the same frequency, thereby constructing a multidimensional data matrix in each circular state. Each row in the multidimensional data matrix represents a wind power parameter at all moments in the preset time period; each column represents the data value corresponding to all types of wind power parameters at the same time. All types of wind power parameters in the multidimensional data matrix are normalized. This application adopts the maximum and minimum normalization method. The implementer can select the normalization method according to the actual situation, and this application does not limit this.

[0044] For various operating states, the average values ​​of all data at the same position in a preset number of multidimensional data matrices are used to form various operating state average data matrices. In this embodiment, the preset number is 1000, and the implementer can adjust it according to the actual situation, and this application does not limit this.

[0045] The second step: based on the differences between various types of wind power parameters and other types in each average data matrix, obtain various types of associated type wind power parameters; perform change point detection on various types of wind power parameters in each average data matrix, analyze the numerical changes of associated type wind power parameters between adjacent change points of various types, and obtain the average correlation coefficient of various types in each average data matrix; merge the number of associated types and the average correlation coefficient of various types to obtain the correlation influence of various types.

[0046] Since the causes of wind power equipment failure are relatively complex, when monitoring the operating status, the status of the generator cannot be judged by only one wind power parameter. It is necessary to comprehensively analyze multiple types of wind power parameters to monitor the status of the generator more comprehensively and accurately. Therefore, this embodiment fuses multiple wind power parameters at the same time, and uses the fused data to train the support vector machine model, so that the trained model can monitor the operating status of the wind power equipment more comprehensively.

[0047] Since there is a certain correlation between different wind power parameters, for example, there is no direct relationship between wind speed and generator failure, but changes in wind speed will affect the change in generator speed, which in turn causes changes in current and power, and changes in current and power may mislead the support vector machine model to judge the normal operation state as a generator failure state. Therefore, when fusing wind power parameters, a comprehensive judgment should be made based on the correlation between different types of wind power parameters.

[0048] Based on this, based on the differences between various types of wind power parameters and other types in each average data matrix, various types of associated type wind power parameters are obtained, specifically:

[0049] The distance metric between any two types of wind power parameters in each average data matrix is ​​calculated and normalized. When the distance metric is greater than a preset correlation threshold, it is determined that there is correlation between the any two types of wind power parameters; otherwise, there is no correlation.

[0050] In each average data matrix, all types of wind power parameters associated with each type are taken as the associated type wind power parameters of each type.

[0051] In this embodiment, the distance measurement between the two types of wind power parameters adopts DTW distance; the normalization method adopts the maximum and minimum method; the association threshold value is 0.8; the implementer can adjust it according to the actual situation, and this application does not impose any restrictions on this.

[0052] Perform change point detection on various types of wind power parameters in each average data matrix, analyze the numerical changes of wind power parameters of associated types between adjacent change points of various types, and obtain the average correlation coefficients of various types in each average data matrix:

[0053] For each average data matrix, change point detection is performed on wind power parameters of various types at all times to obtain a number of trend change points; wherein, this embodiment adopts the CPD (Change point detection) change point detection algorithm. In addition, the schematic diagram of the CPD change point detection algorithm is as follows: Figure 2 As shown; in the figure, the horizontal axis is time, the vertical axis is the normalized gear box temperature, and the dotted line represents the position of the detected trend change point.

[0054] The correlation coefficient of the kth trend change point of the i-th type is recorded as , its formula form is: ;in, Indicates the number of associated types of the i-th type; , They represent the k+1th trend change point and the kth trend change point of the i-th type respectively; , Respectively represent the rth association type of the i-th type , Wind power parameters at the corresponding time; It indicates a preset value greater than zero, with a value of 0.01. The implementer can adjust it according to the actual situation, and this application does not impose any restrictions on this.

[0055] The average level of the correlation coefficients of all trend change points of various types is taken as the average correlation coefficient of various types. In this embodiment, the average level of multiple variables is measured by the average value.

[0056] It should be understood that when The value is smaller, When the value of is large, it means that the rth associated type wind power parameter with the i-th type has caused a large change in the i-th type wind power parameter with a small change, which further indicates that the i-th type wind power parameter is more affected by the associated type wind power parameter at the k-th trend change point. The larger the value of , the greater the influence of the i-th type wind power parameters on the k-th trend change point by the associated type wind power parameters. Furthermore, the greater the correlation of the i-th type wind power parameters at all trend change points, the greater the average correlation coefficient of the i-th type.

[0057] Further, the association influence of each type is obtained according to the number of association types of each type and the average association coefficient. In this embodiment, the product of the number of association types of each type and the average association coefficient is used as the association influence of each type; in some other embodiments, the number of association types of each type is used as the exponent of an exponential function with a natural constant as the base, and the product of the calculation result of the exponential function and the average association coefficient is used as the association influence of each type.

[0058] It should be understood that the greater the number of association types of various types and the greater the average association coefficient, the greater the association influence of various types of wind power parameters.

[0059] The third step: According to the various types of differences between the average data matrices of different operating states, various types of fault representation values ​​are obtained; according to the various types of fault representation values ​​and the associated influence, all wind power parameters at each moment in each multidimensional data matrix are integrated.

[0060] Wind power parameters with stronger correlation may not accurately reflect the operating status, because the fluctuations of these parameters may be caused by faults or the result of changes in other factors. Wind power parameters with weaker correlation may be more able to independently reflect the operating status of the equipment, because their fluctuations are not easily affected by changes in other parameters. When these parameters with strong independence change, it is more likely to be caused by equipment failure rather than other factors. Therefore, using these parameters with weaker correlation to monitor the operating status of the equipment can improve the accuracy of operating status monitoring.

[0061] Therefore, when monitoring the operating status of wind power equipment, each type of wind power parameter has a different degree of importance. So when integrating all types of wind power parameters, different weights should be assigned to each type of wind power parameter according to the importance of the data.

[0062] The importance of each type of wind power parameter is not only related to the correlation influence, but also to the ability of each type of wind power parameter to represent the fault state. If the data of a certain type of wind power parameter under different operating conditions are very different, it means that this type of wind power parameter has a stronger ability to represent the fault state.

[0063] Based on the above characteristics, various types of fault representation values ​​are obtained according to the differences between the average data matrices of different operating states:

[0064] The distances of various types of wind power parameters between any two operating states are obtained, and the average levels of all pairwise combinations of the distances are taken as various types of fault representation values.

[0065] In this embodiment, the distances between various types of wind power parameters in different operating states are measured using Euclidean distances; and the average levels between multiple distances are measured using average values.

[0066] It should be understood that the greater the distance between various types of wind power parameters between different operating states, the greater the ability of this type of wind power parameter to represent faults.

[0067] The correlation influence of a type of wind power parameter reflects the correlation characteristics between it and other types of wind power parameters, while the fault indication value reflects the difference characteristics of this type of wind power parameter under different operating conditions. The smaller the correlation influence of a type of wind power parameter and the larger the fault indication value, the more important this type of wind power parameter is in the fault monitoring task.

[0068] Based on this, various types of weight values ​​are obtained according to various types of fault representation values ​​and associated influences: the normalized value of the ratio of various types of fault representation values ​​to the associated influences is used as the weight value of various types. In order to avoid the situation where the denominator is zero, a preset value needs to be added to the denominator, and the preset value is 0.01; wherein, the normalization method adopts the sigmoid function.

[0069] According to the weight values ​​of various types at each moment in each multidimensional data matrix, all wind power parameters at each moment in each multidimensional data matrix are fused.

[0070] For each moment in each multidimensional data matrix, the product of each type of wind power parameter and its corresponding weight value is calculated, and the cumulative sum of the products of all types of wind power parameters is used as the fusion result of the wind power parameters at each moment in each multidimensional data matrix.

[0071] Among them, the flow chart of obtaining fusion results is as follows: Figure 3 shown.

[0072] The fourth step: train the neural network according to the fusion results of the wind power parameters at each moment in each multidimensional data matrix and the corresponding operating status; monitor the current operating status of the wind power equipment based on the trained neural network.

[0073] The fusion results of the wind power parameters at each moment in the multidimensional data matrix under each operating state are calculated respectively to obtain a fusion data set composed of fusion data. The fusion results in the fusion data set are marked according to the operating state, and the marking content is normal operation, generator failure, blade failure, generator and blade failure. The marked fusion data set is used as the input of the support vector machine model, the support vector machine model is trained, and the trained support vector machine model is used to classify the operating state of the wind power equipment. The training of the support vector machine model is a well-known technology, and the specific process is not repeated here.

[0074] Monitor the various parameters of the wind power equipment in real time to obtain the various parameters at the current moment. Use the weight parameters obtained in the above steps to calculate the fused data at the current moment, and classify the fused data at the current moment through the trained support vector machine model to obtain the operating status of the wind power equipment at the current moment.

[0075] Based on the same inventive concept as the above method, an embodiment of the present application also provides an AI-based multi-parameter monitoring system for wind power equipment, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, it implements the steps of any one of the above-mentioned AI-based multi-parameter monitoring methods for wind power equipment.

[0076] In summary, the embodiments of the present application establish a multidimensional data matrix by analyzing all types of wind power parameters at all times under various operating conditions of wind power equipment, which helps to comprehensively and deeply analyze and understand the complex relationships and patterns of wind power parameters; further obtaining an average data matrix helps to comprehensively analyze various types of data under the same operating state; according to the differences between different types of wind power parameters in the average data matrix, various types of associated wind power parameters are obtained, and based on the correlation characteristics between different types of wind power parameters, an average correlation coefficient is constructed. By comprehensively analyzing the correlation between different types of data, it is evaluated whether these data are affected by mutual correlation, thereby affecting the accuracy of operating status monitoring; further constructing The correlation influence degree helps to understand the relationship between various data types and determine how they jointly affect the accuracy of wind power equipment operating status monitoring; based on the various types of differences between the average data matrices of different operating states, a fault representation value is constructed, which has the beneficial effect of measuring the changing characteristics of various types of wind power parameters under different operating states, and is used to characterize the degree to which the data represents different operating states; based on the correlation influence degree and the fault representation value, the weight value of each type of wind power parameter is calculated, and by assigning different weights to each type of wind power parameter and fusing the data at each moment, it can more accurately reflect the operating status of the wind power equipment, and then use these fused data to train the neural network model, thereby significantly improving the accuracy of operating status monitoring.

[0077] The flowchart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to the embodiment of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. In some alternative implementations, the function marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two continuous boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. In the description corresponding to the flowchart and the block diagram in the accompanying drawings, the operations or steps corresponding to different boxes can also occur in a different order from the order disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two continuous operations or steps can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. Each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented by a dedicated hardware-based system that performs the specified functions or actions, or may be implemented by a combination of dedicated hardware and computer instructions.

[0078] For those skilled in the art, it is obvious that the present application is not limited to the details of the above exemplary embodiments, and the present application can be implemented in other specific forms without departing from the basic features of the present application. Therefore, no matter from which point of view, the above embodiments of the present application should be regarded as exemplary and non-restrictive; the technical solutions recorded in the above embodiments are modified, or some of the technical features are replaced by equivalents, which does not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A wind power equipment multi-parameter monitoring method based on AI, characterized in that: The method comprises the following steps: According to all types of wind power parameters at all times in a preset time period under various operating conditions of the wind power equipment, a multidimensional data matrix under various operating conditions is obtained; for various operating conditions, the average values ​​of all data at the same position of a preset number of multidimensional data matrices are used to form an average data matrix for various operating conditions; Based on the differences between various types of wind power parameters and other types in each average data matrix, various types of associated type wind power parameters are obtained; change point detection is performed on various types of wind power parameters in each average data matrix, and the numerical changes of the associated type wind power parameters between adjacent change points of various types are analyzed to obtain the average association coefficients of various types in each average data matrix; the number of associated types and the average association coefficients of various types are integrated to obtain the association influence of various types; According to the various types of differences between the average data matrices of different operating states, various types of fault representation values ​​are obtained; according to the various types of fault representation values ​​and the associated influence, all wind power parameters at each moment in each multidimensional data matrix are integrated; According to the fusion results of wind power parameters at each moment in each multidimensional data matrix and the corresponding operating status, a neural network is trained; according to the trained neural network, the operating status of the current wind power equipment is monitored; The various types of associated wind power parameters are obtained, specifically: Based on the distance metric between any two types of wind power parameters in each average data matrix, determining whether there is a correlation between the any two types of wind power parameters; All types of wind power parameters associated with various types are used as associated type wind power parameters of various types; The specific process of determining whether there is a correlation between any two types of wind power parameters is as follows: Calculate the distance metric between any two types of wind power parameters in each average data matrix and perform normalization processing, and when the distance metric is greater than a preset correlation threshold, determine that there is correlation between the any two types of wind power parameters; otherwise, there is no correlation; The obtained average correlation coefficients of various types in each average data matrix include: For each average data matrix, change point detection is performed on wind power parameters of various types at all times to obtain several trend change points; The correlation coefficient of the kth trend change point of the i-th type is recorded as , its formula form is: ;in, Indicates the number of associated types of the i-th type; , They represent the k+1th trend change point and the kth trend change point of the i-th type respectively; , Respectively represent the rth association type of the i-th type , Wind power parameters at the corresponding time; Indicates a preset value greater than zero; The average level of the correlation coefficients of all trend change points of various types is taken as the average correlation coefficient of various types.

2. The AI-based wind power equipment multi-parameter monitoring method according to claim 1, characterized in that: The multidimensional data matrix obtained under various operating conditions is specifically: Each row in the multidimensional data matrix represents the wind power parameters of one type at all times within a preset time period; each column represents the wind power parameters of all types at the same time.

3. The AI-based wind power equipment multi-parameter monitoring method according to claim 1, characterized in that: The steps of obtaining various types of fault representation values ​​are as follows: The distances of various types of wind power parameters between any two operating states are obtained, and the average levels of all pairwise combinations of the distances are taken as various types of fault representation values.

4. The AI-based wind power equipment multi-parameter monitoring method according to claim 1, characterized in that: The fusion of all wind power parameters at each moment in each multidimensional data matrix is ​​specifically as follows: The normalized value of the ratio of each type of fault representation value to the associated impact degree is used as the weight value of each type; For each moment in each multidimensional data matrix, the product of each type of wind power parameter and its corresponding weight value is calculated, and the cumulative sum of the products of all types of wind power parameters is used as the fusion result of the wind power parameters at each moment in each multidimensional data matrix.

5. The AI-based wind power equipment multi-parameter monitoring method according to claim 1, characterized in that: The training neural network is specifically as follows: The fusion results of wind power parameters at each moment in the multidimensional data matrix under each operating state are calculated respectively, and the fusion results are marked according to the corresponding operating state. All marked fusion results are used as input of the neural network model to train the neural network model.

6. The AI-based wind power equipment multi-parameter monitoring method according to claim 1, characterized in that: The process of monitoring the current operating status of the wind power equipment includes: The result of fusing all wind power parameters at the current moment is used as the input of the trained neural network model to output the monitoring result of the current wind power equipment operating status.

7. An AI-based wind power equipment multi-parameter monitoring system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.