Wind turbine generator fault diagnosis method and system based on sensor network

Abstract

The invention discloses a wind turbine generator fault diagnosis method and system based on a sensor network, and relates to the field of wind turbine generator fault prediction.The method comprises the steps that specifications and operation data of previous wind turbine generators are obtained, and the operation data of the previous wind turbine generators are analyzed and processed to obtain a wind turbine generator database; constructing a wind turbine generator model and a signal fluctuation curve according to the wind turbine generator database, obtaining the specification of the current wind turbine generator, and carrying out sensor deployment and operation in combination with the wind turbine generator model; acquiring primary sensor data, and configuring an operation strategy for the sensor by adopting an improved tabu search algorithm in combination with the standard signal curve; and implementing a sensor configuration operation strategy and acquiring real-time data of the sensor to obtain a fault diagnosis result of the wind generating set. The improved tabu search algorithm is adopted to optimize and adjust the operation of the sensor, so that the problem of influence on the operation of the wind turbine generator caused by excessive installation and operation of the external sensor is solved, and the fault diagnosis accuracy of the wind turbine generator can be improved.

Description

Technical Field

[0001] The present invention relates to the field of wind turbine fault prediction, and in particular to a wind turbine fault diagnosis method and system based on a sensor network. Background Art

[0002] A wind turbine is a complex system consisting of electrical, mechanical, and automatic control components, and is greatly affected by environmental conditions. In order to prevent major accidents and reduce the failure rate of wind turbines, it is common practice in the industry to conduct big data analysis on the data generated during the operation of wind turbines, to issue early warnings on the status of the wind turbines, and to diagnose faults through data analysis.

[0003] At present, the industry mainly uses SCADA data. Through combing and analysis, SCADA data is cleaned. Missing data can be obtained through fitting methods. Based on the above methods, the quality of data acquisition is improved. Due to the imperfect SCADA of early wind turbines, the quality of information obtained is not high, which makes the accuracy of data analysis low and difficult to guide production. Therefore, it is one of the development trends of the industry to use external data acquisition cards to assist SCADA data, improve data quality, and improve the accuracy of data analysis.

[0004] In order to solve these problems, a wind turbine fault diagnosis method and system based on a sensor network is urgently needed. Summary of the invention

[0005] To solve the above problems, the present application proposes a wind turbine fault diagnosis method and system based on a sensor network, comprising the following steps:

[0006] S1. Obtaining past specifications and operation data of wind turbines, analyzing and processing past operation data of wind turbines to obtain a wind turbine database, wherein the wind turbine database includes a pitch unit database, a current conversion unit database, and a yaw unit database;

[0007] S2. construct a wind turbine model and a signal fluctuation curve according to the wind turbine database;

[0008] S3, obtaining the current wind turbine specifications and combining them with the wind turbine model to deploy and operate sensors;

[0009] S4, obtaining sensor data once and using an improved taboo search algorithm to configure an operating strategy for the sensor in combination with a standard signal curve;

[0010] S5. Implement the sensor configuration operation strategy and obtain the real-time data of the sensor to obtain the fault diagnosis result of the wind turbine generator set.

[0011] Preferably, the specific contents of constructing the wind turbine model and the signal fluctuation curve according to the wind turbine database are:

[0012] The wind turbine model includes a variable pitch model, a variable flow model, and a yaw model;

[0013] Analyze the wind turbine specifications to obtain wind turbine structure information and past sensor installation locations;

[0014] Generate a three-dimensional module of the wind turbine according to the proportion of the structural information of the wind turbine, and disassemble the three-dimensional module of the wind turbine to obtain a variable pitch sub-model, a variable flow sub-model, and a yaw sub-model;

[0015] The sensor nodes are obtained by counting the past sensor installation sites and marked on the corresponding propeller sub-model, variable flow sub-model and yaw sub-model. The sensor nodes include important nodes and ordinary nodes.

[0016] Match the operation data with the sensor nodes and establish the operation curve of the corresponding sensor nodes with time as the horizontal axis;

[0017] The features of the operation curve in different time periods are extracted to obtain stable features and abnormal features;

[0018] According to the stable characteristics, the running curve is intercepted and divided to obtain the stable curve, and the stable curve is fitted to obtain the standard signal fluctuation curve and the standard fluctuation range value;

[0019] The operation curve is intercepted and divided according to abnormal characteristics to obtain an abnormal curve, and the abnormal characteristics correspond to different wind turbine faults.

[0020] Preferably, the sensor nodes include a sensor influence relationship, and a first influence factor is set between the sensor nodes according to the sensor influence relationship;

[0021] The sensor node is provided with a sensor node coordinate, and the sensor node coordinate takes the bottom of one end of the wind turbine model as the origin coordinate.

[0022] Preferably, the specific contents of obtaining the current wind turbine specifications and combining the wind turbine model to deploy sensors include:

[0023] Obtaining current wind turbine specifications, obtaining current wind turbine structure information according to the current wind turbine specifications, and generating a current wind turbine three-dimensional model according to the current wind turbine structure information;

[0024] The current wind turbine three-dimensional model is disassembled to obtain the current variable pitch sub-model, the current variable flow sub-model, and the current yaw sub-model;

[0025] According to the current pitch sub-model, the current flow sub-model and the current yaw sub-model, the corresponding current sensor nodes and matching signal fluctuation curves are matched in the pitch model, the flow model and the yaw model respectively;

[0026] Sensors are deployed and operated for the current wind turbine according to the current sensor nodes.

[0027] Preferably, the specific contents of acquiring sensor data once and configuring the sensor operation strategy using an improved taboo search algorithm in combination with the standard signal curve also include:

[0028] S401, setting an objective function according to the coverage and quantity of sensors;

[0029] S402, the current sensor node is an initial sensor network layout and the initial sensor network layout is used as a starting point;

[0030] S403, changing the positions of the sensor nodes starting from the starting point to generate a secondary layout, i.e., a neighborhood solution;

[0031] S404, calculating the objective function values ​​corresponding to the secondary layouts, sorting the objective function values ​​and selecting the secondary layout with the highest order as the current target layout;

[0032] S405, a target range is preset, and the current target layout and the layouts within the adjacent target range are added to the taboo table;

[0033] S406, a target number of iterations is preset, and S403 to S405 are repeated until the target number of iterations is reached;

[0034] S407: Select the optimal sensor network layout from the taboo table as the target result, and configure the sensor operation strategy according to the target result.

[0035] Preferably, the objective function is:

[0036]

[0037] Among them, W is the objective function value, α is the distance conversion coefficient, β is the energy consumption conversion coefficient, and X max is the maximum value of node coordinates in the initial sensor network layout, x max is the maximum value of node coordinates in the secondary layout, is the coordinate x i The energy consumption value of the sensor corresponding to the node, is the coordinate x max The energy consumption value of the sensor corresponding to the node.

[0038] Preferably, in the process of changing the position of the sensor node from the starting point to generate the secondary layout, i.e., the neighborhood solution, the specific manner of changing the position of the sensor node is:

[0039] A target one-time operation time is preset, and the one-time sensor data is the sensor operation data within the one-time operation time;

[0040] The sensor operation data at different nodes are used to establish the current signal fluctuation curve with time as the coordinate;

[0041] Matching the current signal fluctuation curve with the matching signal fluctuation curve to obtain the current standard signal fluctuation curve;

[0042] Extract the current signal fluctuation curve characteristics. If the current signal fluctuation curve characteristics meet the stable characteristics of the current standard signal fluctuation curve, retain the important nodes to change the position of the sensor node.

[0043] If the current signal fluctuation curve characteristics meet the abnormal characteristics of the current standard signal fluctuation curve, the important nodes and the nodes corresponding to the abnormal characteristics are retained to obtain a change in the position of the sensor node;

[0044] Calculate the offset of the current signal fluctuation curve corresponding to the abnormal feature, and multiply the offset and the first impact factor between the node corresponding to the abnormal feature and other sensor nodes to obtain the node emphasis value;

[0045] A node emphasis value threshold is preset. If the node emphasis value exceeds the node emphasis value threshold, the node exceeding the node emphasis value threshold is selected as the position of the secondary changed sensor node.

[0046] A wind turbine fault diagnosis system based on a sensor network, comprising:

[0047] Data acquisition unit: obtains past wind turbine specifications and operation data, and analyzes and processes past wind turbine operation data to obtain a wind turbine database, wherein the wind turbine database includes a pitch unit database, a current conversion unit database, and a yaw unit database;

[0048] Model building unit: building wind turbine model and signal fluctuation curve according to wind turbine database;

[0049] Strategy operation unit: obtain the current wind turbine specifications and combine them with the wind turbine model to deploy and operate the sensor, obtain the primary sensor data and use the improved taboo search algorithm to configure the sensor operation strategy in combination with the standard signal curve;

[0050] Result judgment unit: implements the sensor configuration operation strategy and obtains the real-time data of the sensor to obtain the fault diagnosis result of the wind turbine generator set.

[0051] An electronic device, characterized in that it includes a memory and a processor, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the content of the wind turbine fault diagnosis method based on the sensor network is implemented.

[0052] A storage medium, characterized in that the storage medium stores computer executable instructions, and when the computer executable instructions are loaded and executed by a processor, the content of the wind turbine fault diagnosis method based on a sensor network is implemented.

[0053] In summary, the wind turbine fault diagnosis method and system based on sensor network of the present invention, compared with the traditional technology, adopts the improved taboo search algorithm to optimize the sensor operation, which improves the problem of excessive installation and operation of external sensors affecting the operation of wind turbines and improves the accuracy of wind turbine fault diagnosis. The present invention can reduce the fault location and processing time. At the same time, according to the in-depth analysis of the fault, operation and maintenance technical measures can be formulated to reduce the overall failure rate and save spare parts procurement expenses.

[0054] The technical method of the present invention is further described in detail below through the drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A step diagram of a wind turbine fault diagnosis method based on a sensor network according to the present invention;

[0056] Figure 2 This is a module diagram of a wind turbine fault diagnosis system based on a sensor network according to the present invention. DETAILED DESCRIPTION

[0057] The technical method of the present invention is further described below by means of the accompanying drawings and embodiments. It should be noted that unless otherwise specifically stated, the relative arrangement of the components and steps, the numerical expressions and the numerical values ​​described in these embodiments do not limit the scope of the present application.

[0058] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present application, its application, or uses.

[0059] Technologies, systems, and devices known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, systems, and devices should be considered part of the specification.

[0060] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0061] Unless otherwise defined, technical or scientific terms used in the present invention shall have the common meanings understood by one having ordinary skills in the field to which the present invention belongs.

[0062] The present invention provides a wind turbine fault diagnosis method and system based on a sensor network, S1, obtaining past wind turbine specifications and operation data, analyzing and processing the past wind turbine operation data to obtain a wind turbine database, wherein the wind turbine database includes a pitch unit database, a current conversion unit database, and a yaw unit database;

[0063] S2. construct a wind turbine model and a signal fluctuation curve according to the wind turbine database;

[0064] Preferably, the specific contents of constructing the wind turbine model and the signal fluctuation curve according to the wind turbine database are:

[0065] The wind turbine model includes a variable pitch model, a variable flow model, and a yaw model;

[0066] Analyze the wind turbine specifications to obtain wind turbine structure information and past sensor installation locations;

[0067] Generate a three-dimensional module of the wind turbine according to the proportion of the structural information of the wind turbine, and disassemble the three-dimensional module of the wind turbine to obtain a variable pitch sub-model, a variable flow sub-model, and a yaw sub-model;

[0068] The sensor nodes are obtained by counting the past sensor installation sites and marked on the corresponding propeller sub-model, variable flow sub-model and yaw sub-model. The sensor nodes include important nodes and ordinary nodes.

[0069] Match the operation data with the sensor nodes and establish the operation curve of the corresponding sensor nodes with time as the horizontal axis;

[0070] The features of the operation curve in different time periods are extracted to obtain stable features and abnormal features;

[0071] According to the stable characteristics, the running curve is intercepted and divided to obtain the stable curve, and the stable curve is fitted to obtain the standard signal fluctuation curve and the standard fluctuation range value;

[0072] The operation curve is intercepted and divided according to abnormal characteristics to obtain an abnormal curve, and the abnormal characteristics correspond to different wind turbine faults.

[0073] Preferably, the sensor nodes include a sensor influence relationship, and a first influence factor is set between the sensor nodes according to the sensor influence relationship;

[0074] The sensor node is provided with a sensor node coordinate, and the sensor node coordinate takes the bottom of one end of the wind turbine model as the origin coordinate.

[0075] S3, obtaining the current wind turbine specifications and combining them with the wind turbine model to deploy and operate sensors;

[0076] Preferably, the specific contents of obtaining the current wind turbine specifications and combining the wind turbine model to deploy sensors include:

[0077] Obtaining current wind turbine specifications, obtaining current wind turbine structure information according to the current wind turbine specifications, and generating a current wind turbine three-dimensional model according to the current wind turbine structure information;

[0078] The current wind turbine three-dimensional model is disassembled to obtain the current variable pitch sub-model, the current variable flow sub-model, and the current yaw sub-model;

[0079] According to the current pitch sub-model, the current flow sub-model and the current yaw sub-model, the corresponding current sensor nodes and matching signal fluctuation curves are matched in the pitch model, the flow model and the yaw model respectively;

[0080] Sensors are deployed and operated for the current wind turbine according to the current sensor nodes.

[0081] S4, obtaining sensor data once and using an improved taboo search algorithm to configure an operating strategy for the sensor in combination with a standard signal curve;

[0082] Preferably, the specific contents of acquiring sensor data once and configuring the sensor operation strategy using an improved taboo search algorithm in combination with the standard signal curve also include:

[0083] S401, setting an objective function according to the coverage and quantity of sensors;

[0084] S402, the current sensor node is an initial sensor network layout and the initial sensor network layout is used as a starting point;

[0085] S403, changing the positions of the sensor nodes starting from the starting point to generate a secondary layout, i.e., a neighborhood solution;

[0086] S404, calculating the objective function values ​​corresponding to the secondary layouts, sorting the objective function values ​​and selecting the secondary layout with the highest order as the current target layout;

[0087] S405, a target range is preset, and the current target layout and layouts within the target range adjacent to it are added to a taboo table to avoid repeated access to these layouts in future searches;

[0088] S406, a target number of iterations is preset, and S403 to S405 are repeated until the target number of iterations is reached;

[0089] S407: Select the optimal sensor network layout from the taboo table as the target result, and configure the sensor operation strategy according to the target result.

[0090] Preferably, the objective function is:

[0091]

[0092] Among them, W is the objective function value, α is the distance conversion coefficient, β is the energy consumption conversion coefficient, and X max is the maximum value of node coordinates in the initial sensor network layout, x max is the maximum value of node coordinates in the secondary layout, is the coordinate x i The energy consumption value of the sensor corresponding to the node, is the coordinate x max The energy consumption value of the sensor corresponding to the node.

[0093] Preferably, in the process of changing the position of the sensor node from the starting point to generate the secondary layout, i.e., the neighborhood solution, the specific manner of changing the position of the sensor node is:

[0094] A target one-time operation time is preset, and the one-time sensor data is the sensor operation data within the one-time operation time;

[0095] The sensor operation data at different nodes are used to establish the current signal fluctuation curve with time as the coordinate;

[0096] Matching the current signal fluctuation curve with the matching signal fluctuation curve to obtain the current standard signal fluctuation curve;

[0097] Extract the current signal fluctuation curve characteristics. If the current signal fluctuation curve characteristics meet the stable characteristics of the current standard signal fluctuation curve, retain the important nodes to change the position of the sensor node.

[0098] If the current signal fluctuation curve characteristics meet the abnormal characteristics of the current standard signal fluctuation curve, the important nodes and the nodes corresponding to the abnormal characteristics are retained to obtain a change in the position of the sensor node;

[0099] Calculate the offset of the current signal fluctuation curve corresponding to the abnormal feature, and multiply the offset and the first impact factor between the node corresponding to the abnormal feature and other sensor nodes to obtain the node emphasis value;

[0100] A node emphasis value threshold is preset. If the node emphasis value exceeds the node emphasis value threshold, the node exceeding the node emphasis value threshold is selected as the position of the secondary changed sensor node.

[0101] S5. Implement the sensor configuration operation strategy and obtain the real-time data of the sensor to obtain the fault diagnosis result of the wind turbine generator set.

[0102] A wind turbine fault diagnosis system based on a sensor network, comprising:

[0103] Data acquisition unit: obtains past wind turbine specifications and operation data, and analyzes and processes past wind turbine operation data to obtain a wind turbine database, wherein the wind turbine database includes a pitch unit database, a current conversion unit database, and a yaw unit database;

[0104] Model building unit: building wind turbine model and signal fluctuation curve according to wind turbine database;

[0105] Strategy operation unit: obtain the current wind turbine specifications and combine them with the wind turbine model to deploy and operate the sensor, obtain the primary sensor data and use the improved taboo search algorithm to configure the sensor operation strategy in combination with the standard signal curve;

[0106] Result judgment unit: implements the sensor configuration operation strategy and obtains the real-time data of the sensor to obtain the fault diagnosis result of the wind turbine generator set.

[0107] An electronic device, characterized in that it includes a memory and a processor, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the content of the wind turbine fault diagnosis method based on the sensor network is implemented.

[0108] A storage medium, characterized in that the storage medium stores computer executable instructions, and when the computer executable instructions are loaded and executed by a processor, the content of the wind turbine fault diagnosis method based on a sensor network is implemented.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical method of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical method to deviate from the spirit and scope of the technical method of the present invention.

Claims

1. A wind turbine fault diagnosis method based on a sensor network, characterized in that: The following steps are involved: S1. Obtaining past specifications and operation data of wind turbines, analyzing and processing past operation data of wind turbines to obtain a wind turbine database, wherein the wind turbine database includes a pitch unit database, a current conversion unit database, and a yaw unit database; S2. construct a wind turbine model and a signal fluctuation curve according to the wind turbine database; S3, obtaining the current wind turbine specifications and combining them with the wind turbine model to deploy and operate sensors; S4, obtaining sensor data once and using an improved taboo search algorithm to configure an operating strategy for the sensor in combination with a standard signal curve; S5. Implement the sensor configuration operation strategy and obtain the real-time data of the sensor to obtain the fault diagnosis result of the wind turbine generator set.

2. A wind turbine fault diagnosis method based on a sensor network according to claim 1, characterized in that: The specific contents of constructing the wind turbine model and signal fluctuation curve based on the wind turbine database are as follows: The wind turbine model includes a variable pitch model, a variable flow model, and a yaw model; Analyze the wind turbine specifications to obtain wind turbine structure information and past sensor installation locations; Generate a three-dimensional module of the wind turbine according to the proportion of the structural information of the wind turbine, and disassemble the three-dimensional module of the wind turbine to obtain a variable pitch sub-model, a variable flow sub-model, and a yaw sub-model; The sensor nodes are obtained by counting the past sensor installation sites and marked on the corresponding propeller sub-model, variable flow sub-model and yaw sub-model. The sensor nodes include important nodes and ordinary nodes. Match the operation data with the sensor nodes and establish the operation curve of the corresponding sensor nodes with time as the horizontal axis; The features of the operation curve in different time periods are extracted to obtain stable features and abnormal features; According to the stable characteristics, the running curve is intercepted and divided to obtain the stable curve, and the stable curve is fitted to obtain the standard signal fluctuation curve and the standard fluctuation range value; The operation curve is intercepted and divided according to abnormal characteristics to obtain an abnormal curve, and the abnormal characteristics correspond to different wind turbine faults.

3. A wind turbine fault diagnosis method based on a sensor network according to claim 2, characterized in that: The sensor nodes include a sensor influence relationship, and a first influence factor is set between the sensor nodes according to the sensor influence relationship; The sensor node is provided with a sensor node coordinate, and the sensor node coordinate takes the bottom of one end of the wind turbine model as the origin coordinate.

4. A wind turbine fault diagnosis method based on a sensor network according to claim 3, characterized in that: The specific contents of obtaining the current wind turbine specifications and combining them with the wind turbine model for sensor deployment include: Obtaining current wind turbine specifications, obtaining current wind turbine structure information according to the current wind turbine specifications, and generating a current wind turbine three-dimensional model according to the current wind turbine structure information; The current wind turbine three-dimensional model is disassembled to obtain the current variable pitch sub-model, the current variable flow sub-model, and the current yaw sub-model; According to the current pitch sub-model, the current flow sub-model and the current yaw sub-model, the corresponding current sensor nodes and matching signal fluctuation curves are matched in the pitch model, the flow model and the yaw model respectively; Sensors are deployed and operated for the current wind turbine according to the current sensor nodes.

5. A wind turbine fault diagnosis method based on a sensor network according to claim 4, characterized in that: The specific contents of obtaining sensor data once and using the improved taboo search algorithm to configure the sensor operation strategy in combination with the standard signal curve also include: S401, setting an objective function according to the coverage and quantity of sensors; S402, the current sensor node is an initial sensor network layout and the initial sensor network layout is used as a starting point; S403, changing the positions of the sensor nodes starting from the starting point to generate a secondary layout, i.e., a neighborhood solution; S404, calculating the objective function values ​​corresponding to the secondary layouts, sorting the objective function values ​​and selecting the secondary layout with the highest order as the current target layout; S405, a target range is preset, and the current target layout and the layouts within the adjacent target range are added to the taboo table; S406, a target number of iterations is preset, and S403 to S405 are repeated until the target number of iterations is reached; S407: Select the optimal sensor network layout from the taboo table as the target result, and configure the sensor operation strategy according to the target result.

6. A wind turbine fault diagnosis method based on a sensor network according to claim 5, characterized in that: The objective function is: Among them, W is the objective function value, α is the distance conversion coefficient, β is the energy consumption conversion coefficient, and X max is the maximum value of node coordinates in the initial sensor network layout, x max is the maximum value of node coordinates in the secondary layout, is the coordinate x i The energy consumption value of the sensor corresponding to the node, is the coordinate x max The energy consumption value of the sensor corresponding to the node.

7. A wind turbine fault diagnosis method based on a sensor network according to claim 6, characterized in that: In the process of changing the position of the sensor node from the starting point to generate the secondary layout, i.e. the neighborhood solution, the specific way to change the position of the sensor node is as follows: A target one-time operation time is preset, and the one-time sensor data is the sensor operation data within the one-time operation time; The sensor operation data at different nodes are used to establish the current signal fluctuation curve with time as the coordinate; Matching the current signal fluctuation curve with the matching signal fluctuation curve to obtain the current standard signal fluctuation curve; Extract the current signal fluctuation curve characteristics. If the current signal fluctuation curve characteristics meet the stable characteristics of the current standard signal fluctuation curve, retain the important nodes to change the position of the sensor node. If the current signal fluctuation curve characteristics meet the abnormal characteristics of the current standard signal fluctuation curve, the important nodes and the nodes corresponding to the abnormal characteristics are retained to obtain a change in the position of the sensor node; Calculate the offset of the current signal fluctuation curve corresponding to the abnormal feature, and multiply the offset and the first impact factor between the node corresponding to the abnormal feature and other sensor nodes to obtain the node emphasis value; A node emphasis value threshold is preset. If the node emphasis value exceeds the node emphasis value threshold, the node exceeding the node emphasis value threshold is selected as the position of the secondary changed sensor node.

8. A wind turbine fault diagnosis system based on a sensor network, characterized in that: include: Data acquisition unit: obtains past specifications and operation data of wind turbines, and analyzes and processes past operation data of wind turbines to obtain a wind turbine database, wherein the wind turbine database includes a pitch unit database, a current conversion unit database, and a yaw unit database; Model building unit: building wind turbine model and signal fluctuation curve according to wind turbine database; Strategy operation unit: obtain the current wind turbine specifications and combine them with the wind turbine model to deploy and operate the sensor, obtain the primary sensor data and use the improved taboo search algorithm to configure the sensor operation strategy in combination with the standard signal curve; Result judgment unit: implements the sensor configuration operation strategy and obtains the real-time data of the sensor to obtain the fault diagnosis result of the wind turbine generator set.

9. An electronic device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and when the processor calls the computer program in the memory, the content of the wind turbine fault diagnosis method based on the sensor network as claimed in any one of claims 1 to 7 is implemented.

10. A storage medium, characterized in that: The storage medium stores computer executable instructions, which, when loaded and executed by a processor, implement the contents of the wind turbine fault diagnosis method based on a sensor network as claimed in any one of claims 1 to 7.

Images