A kind of wind power generation equipment operation and maintenance management inspection method

Abstract

The application discloses a kind of wind power generation equipment operation and maintenance management's inspection method, it is related to wind power generation equipment intelligent inspection technical field, the method includes obtaining the basic information of user selection to be inspected wind turbine generator unit and corresponding gear box, output to-be-inspected object determination result, gear box schematic diagram is displayed according to the to-be-inspected object determination result, according to the initial labeling point selected by user, collect axial, flatness, coaxiality deviation, and comprehensively obtain measured deviation value.Combining the associated influence change trend characteristics between historical deviation value and historical wind field characteristic data, error compensation is carried out on measured deviation value to obtain actual deviation value, according to historical detection data, construct misjudgment prediction model, and according to actual deviation value, measure misjudgment rate, and accordingly expand labeling point.Collect on-site picture and gear box operating parameter, form inspection data archives.According to user-selected object and time, output full-process inspection result, the application can improve wind power equipment inspection full-process management efficiency.

Description

Technical Field

[0001] This application relates to the field of intelligent inspection technology for wind power generation equipment, and in particular to an inspection method for the operation and maintenance management of wind power generation equipment. Background Technology

[0002] As a core transmission component of wind turbine generators, the gearbox's operating condition directly affects the generator's power generation efficiency and overall operational safety. Endoscopic inspection is a crucial maintenance method for assessing the gearbox's internal condition and diagnosing faults. Currently, the wind power industry still faces several technical shortcomings in gearbox endoscopic inspection and related maintenance, primarily in the following aspects:

[0003] During the process of selecting corresponding locations for data detection using the marked points on the loaded structural schematic diagram, there is a deviation between the marked points and the inspection locations uploaded by the user, which leads to misjudgments in the detection data results. No further mark point expansion and optimization management has been performed.

[0004] At the data collection and management level, current inspections mostly use paper documents or common electronic spreadsheets such as Excel for information recording, resulting in a low degree of standardization and normalization. Inconsistent recording formats and terminology among different maintenance personnel easily lead to problems such as missing key inspection information and poor data standardization, while also hindering centralized data aggregation, statistics, and in-depth analysis. Various types of inspection information are isolated from each other; basic wind farm information, wind turbine ledgers, results of previous endoscopic inspections, on-site video data, and abnormal defect records are stored in a scattered manner, forming numerous information silos. This makes it difficult for management personnel to comprehensively grasp the historical status and changing trends of individual wind turbines or even the entire gearbox network, hindering effective correlation analysis and trend judgment.

[0005] In the inspection and anomaly identification phases, the standardization of inspection procedures and the determination of anomalies heavily rely on the personal experience of maintenance personnel, lacking standardized operating guidelines and objective and unified evaluation criteria. In actual operations, missed or incorrect inspections are easily caused by human error such as skipping steps or differences in experience. The judgment of anomaly types and severity is highly subjective and inconsistent, making it difficult to achieve accurate and reliable anomaly identification and unable to support intelligent early warning of faults.

[0006] In the data management and report preparation stages, the on-site photos taken by endoscopy lack effective correlation with the examined areas and results, file naming and archiving are chaotic, and retrieval and reuse efficiency is low. Inspection reports require manual compilation and organization of various scattered records and matching with image data, resulting in a large workload, long processing time, inconsistent report formats, and insufficient standardization, leading to overall low operational and maintenance efficiency. Summary of the Invention

[0007] To overcome the shortcomings of the prior art, this application provides an inspection method for the operation and maintenance management of wind power generation equipment.

[0008] This application provides an inspection method for the operation and maintenance management of wind power generation equipment, the method comprising:

[0009] Step S1: Obtain the basic information of the wind turbine generator set and corresponding gearbox selected by the user, output the result of the object to be inspected, display the gearbox schematic diagram according to the result of the object to be inspected, and collect the axial deviation, flatness deviation and coaxiality deviation according to the positional relationship of the initial marked inspection point selected by the user on the gearbox schematic diagram, which belongs to the straight line layout, plane layout and three-dimensional space layout, and obtain the measured deviation value.

[0010] Step S2: Based on the correlation and trend characteristics between historical deviation values ​​and historical wind farm characteristic data in historical detection data, error compensation is performed on the deviation value to be measured to obtain the actual deviation value. Based on historical detection data, a misjudgment prediction model is established. The actual deviation value is input into the misjudgment prediction model for testing to obtain the current misjudgment rate. Based on the current misjudgment rate, the initial marked checkpoints are expanded to obtain expanded marked checkpoints. Based on the initial marked checkpoints and expanded marked checkpoints, on-site images uploaded by users are collected, and preprocessed image information is obtained by combining them.

[0011] Step S3: Based on the initial and expanded checkpoints, collect the operating parameters of the gearbox in the determination results of the object to be inspected, and obtain the preprocessed inspection data. Based on the preprocessed image information and the preprocessed inspection data, obtain the inspection data archive. Based on the inspection data archive and the object and time range selected by the user to generate the report, output the full-process inspection results.

[0012] Preferably, the system obtains the user's identity authentication information when entering the wind power equipment inspection platform, obtains the user's permission verification information based on the identity authentication information, displays the main interface of the wind power equipment inspection platform to the user based on the permission verification information, and outputs the display results of the main interface of the platform.

[0013] The user selects the basic information of the wind turbine generator set to be inspected and the corresponding gearbox from the results displayed on the main interface of the platform, and the result of the inspection object determination is output.

[0014] Based on the gearbox model selected from the results of the test object determination, display the internal structure diagram of the gearbox and output the gearbox schematic diagram.

[0015] Preferably, the user selects the initial annotation checkpoints on the gearbox schematic diagram. If all the initial annotation checkpoints are arranged in a straight line, the straight line arrangement position condition is output.

[0016] Based on the straight layout position conditions, the axial deviation of the gearbox in the determination result of the object to be inspected is collected to obtain the axial deviation value;

[0017] If all checkpoints in the initial checkpoints are arranged in a planar layout, output the planar layout position conditions.

[0018] Based on the planar layout conditions, the flatness deviation of the gearbox in the determination result of the object to be inspected is collected to obtain the flatness deviation value;

[0019] If all the initial annotation checkpoints are arranged in a three-dimensional spatial relationship, output the spatial arrangement position conditions.

[0020] Based on the spatial layout conditions, the coaxiality deviation of the gearbox in the determination result of the object to be inspected is collected to obtain the three-dimensional spatial deviation value.

[0021] The axial deviation value, flatness deviation value, and three-dimensional spatial deviation value are combined to form the deviation value to be measured.

[0022] Preferably, the temperature, humidity, wind speed, and vibration data of the environment to which the gearbox belongs are collected from the determination results of the object to be tested, so as to obtain the characteristic data of the wind farm to be tested;

[0023] Historical detection data is obtained, and the correlation influence coefficient between historical deviation values ​​and historical wind farm characteristic data is obtained based on the correlation influence trend characteristics between the historical deviation values ​​and historical wind farm characteristic data in the historical detection data.

[0024] Preferably, the pre-compensation deviation value is obtained based on the characteristic data of the wind farm to be tested and the correlation influence coefficient;

[0025] The actual deviation value is obtained by subtracting the measured deviation value from the pre-compensation deviation value.

[0026] Based on the historical deviation values, the historical misjudgment rate in the historical detection data, and the historical gearbox operation data in the historical detection data, a misjudgment prediction model is constructed.

[0027] The operating parameters of the gearbox in the determination result of the object to be inspected are collected to obtain the inspection data to be tested. The inspection data to be tested and the actual deviation value are input into the misjudgment prediction model for testing to obtain the current misjudgment rate.

[0028] Preferably, a false judgment threshold is preset. If the current false judgment rate is less than the false judgment threshold, there is no need to expand the initial annotation checkpoints. Based on the initial annotation checkpoints, the on-site images uploaded by the user are collected to obtain the image information to be tested.

[0029] If the current misjudgment rate is greater than or equal to the misjudgment threshold, the initial annotation checkpoint is extended and expanded according to the linear layout position condition to obtain amplified annotation point one; or, the initial annotation checkpoint is expanded diagonally according to the planar layout position condition to obtain amplified annotation point two; or, the initial annotation checkpoint is expanded layer by layer according to the spatial layout position condition to obtain amplified annotation point three. The amplified annotation point one, amplified annotation point two, and amplified annotation point three are combined to form the amplified annotation checkpoint.

[0030] Obtain additional checkpoints selected by the user from the amplified annotation checkpoints, and collect on-site images uploaded by the user based on the additional checkpoints to obtain the second image information to be tested;

[0031] The image information to be tested, part one, and the image information to be tested, part two, are combined to form preprocessed image information.

[0032] Preferably, based on the amplified and marked checkpoints, the operating parameters of the gearbox in the determination result of the object to be inspected are collected to obtain additional inspection data, and the additional inspection data and the inspection data to be tested are combined to form preprocessed inspection data;

[0033] Based on historical detection data, a pre-set abnormal keyword rule base is used to identify anomalies in the pre-processed inspection data and obtain identification results. If the identification result is abnormal and hits the abnormal keyword rule base, an abnormal record is created and the abnormal record information is output.

[0034] The preprocessed image information, preprocessed inspection data, and abnormal record information are stored, synchronized, and their consistency is verified to obtain the inspection data archive.

[0035] The system obtains the object and time range selected by the user for generating the report, selects relevant data from the inspection data archive based on the selected object and time range, generates a standardized professional inspection report, transmits the professional inspection report to the user, and outputs the full-process inspection results.

[0036] Compared with the prior art, the present invention has the following characteristics and beneficial effects:

[0037] By locating initial checkpoints using a gearbox schematic diagram and combining the axial, flatness, and coaxiality deviations collected based on the layout position, the measured deviation value is obtained comprehensively. This avoids the one-sidedness of single deviation detection. Furthermore, traditional techniques fail to consider the positional relationship between marked checkpoints when determining deviations. For user-selected checkpoints, corresponding gearbox operating data is directly measured without further verification to address potential risks of misjudgment, leading to inaccurate final results and reduced efficiency. By leveraging historical data to mine the correlation between deviations and wind farm characteristics, the measured deviation is analyzed for error correction. Compensation corrects detection errors caused by environmental interference. Simultaneously, it tests the current misjudgment rate through a misjudgment prediction model, dynamically expands checkpoints, and improves inspection accuracy. By dynamically adjusting checkpoints based on the misjudgment rate, it reduces misjudgments and omissions, eliminating the need for manual blind expansion, reducing inspection workload, and improving operation and maintenance efficiency. Through a pre-set abnormal keyword rule base, it can automatically identify and classify potential abnormalities from pre-processed inspection data, reducing the risk of missed judgments and improving the consistency and timeliness of abnormality management. Through unified data management, it organically links all information such as wind farms, equipment, inspection items, photos, abnormalities, and reports, breaking down information silos. Attached Figure Description

[0038] Figure 1 This is a flowchart illustrating the steps of an inspection method for the operation and maintenance management of wind power generation equipment, which is the main feature of this embodiment. Detailed Implementation

[0039] The present invention will be further described in detail below with reference to the following embodiments.

[0040] Reference Figure 1 An inspection method for the operation and maintenance management of wind power generation equipment, the method comprising the following steps:

[0041] Step S1: Obtain the basic information of the wind turbine generator set and corresponding gearbox selected by the user, output the result of the object to be inspected, display the gearbox schematic diagram based on the result of the object to be inspected, and collect the axial deviation, flatness deviation and coaxiality deviation based on the positional relationship of the initial marked inspection point selected by the user on the gearbox schematic diagram to the linear layout, planar layout and three-dimensional spatial layout, and obtain the measured deviation value.

[0042] Step S2: Based on the correlation and trend characteristics between historical deviation values ​​and historical wind farm characteristic data in historical detection data, error compensation is performed on the deviation value to be measured to obtain the actual deviation value. Based on historical detection data, a misjudgment prediction model is established. The actual deviation value is input into the misjudgment prediction model for testing to obtain the current misjudgment rate. Based on the current misjudgment rate, the initial marked checkpoints are expanded to obtain expanded marked checkpoints. Based on the initial marked checkpoints and expanded marked checkpoints, on-site images uploaded by users are collected, and preprocessed image information is obtained by combining them.

[0043] Step S3: Based on the initial and expanded checkpoints, collect the operating parameters of the gearbox in the determination results of the object to be inspected, and obtain the preprocessed inspection data. Based on the preprocessed image information and the preprocessed inspection data, obtain the inspection data archive. Based on the inspection data archive and the object and time range selected by the user to generate the report, output the full-process inspection results.

[0044] Specifically, by locating initial checkpoints using a gearbox schematic diagram and combining the axial, flatness, and coaxiality deviations collected based on the layout position, the measured deviation value is obtained comprehensively. This avoids the one-sidedness of single deviation detection. Furthermore, traditional techniques do not consider the positional relationship between marked checkpoints when determining deviations. For user-selected checkpoints, corresponding gearbox operating data is directly tested without further verification to address potential risks of misjudgment, leading to inaccurate final inspection results and reduced efficiency of the entire inspection process. By utilizing historical data to mine the correlation between deviations and wind farm characteristics, the measured deviation is further... Line error compensation corrects detection errors caused by environmental interference. At the same time, it tests the current misjudgment rate through a misjudgment prediction model, dynamically expands checkpoints, and improves inspection accuracy. By dynamically adjusting checkpoints based on the misjudgment rate, it reduces misjudgments and omissions, eliminating the need for manual blind expansion, reducing inspection workload, and improving operation and maintenance efficiency. Through a pre-set abnormal keyword rule base, it can automatically identify and classify potential abnormalities from pre-processed inspection data, reducing the risk of missed judgments and improving the consistency and timeliness of abnormality management. Through unified data management, it organically links all information such as wind farms, equipment, inspection items, photos, abnormalities, and reports, breaking down information silos.

[0045] The specific step S1 includes the following sub-steps:

[0046] Obtain the user's identity authentication information when entering the wind power equipment inspection platform. Based on the identity authentication information, obtain the user's permission verification information. Based on the permission verification information, display the main interface of the wind power equipment inspection platform to the user and output the display results of the main interface.

[0047] Users select the basic information of the wind turbine generator set to be inspected and the corresponding gearbox from the results displayed on the platform's main interface, and the results of the inspection are output.

[0048] Based on the gearbox model selected in the results of the inspection object determination, display the internal structure diagram of the gearbox and output the gearbox schematic diagram.

[0049] Obtain the initial annotation checkpoints selected by the user on the gearbox schematic diagram. If all the initial annotation checkpoints are arranged in a straight line, output the straight line arrangement position condition.

[0050] Based on the linear layout conditions, the axial deviation of the gearbox in the determination results of the object to be inspected is collected to obtain the axial deviation value.

[0051] If all checkpoints in the initial checkpoints are arranged in a planar layout, output the planar layout position conditions.

[0052] Based on the planar layout conditions, the flatness deviation of the gearbox in the determination results of the object to be inspected is collected to obtain the flatness deviation value.

[0053] If all the initial annotation checkpoints are arranged in a three-dimensional spatial positional relationship, output the spatial arrangement position conditions.

[0054] Based on the spatial layout conditions, the coaxiality deviation of the gearbox in the determination results of the object to be inspected is collected to obtain the three-dimensional spatial deviation value.

[0055] The axial deviation value, flatness deviation value, and three-dimensional spatial deviation value are combined to form the deviation value to be measured.

[0056] Specifically, the results displayed on the platform's main interface (this wind power equipment inspection platform includes the following modules: user login and main interface module, wind farm and gearbox management module, gearbox visualization inspection module, anomaly intelligent identification and management module, and intelligent report generation and email sending module. Among them, the user login and main interface module is responsible for user identity authentication and permission management (permission verification) and serves as the integrated main window, i.e., displaying the platform's main interface), and the results of determining the objects to be inspected (based on the wind farm and gearbox management module: used to maintain the basic information database of wind farms, wind turbine generators, and their gearboxes (such as manufacturer, model, structure type, lubricating oil information, etc., i.e., users select or maintain the wind farms, wind turbine generators, and gearboxes to be inspected in the wind farm and gearbox management module). Based on the basic information of the gearbox, the object to be inspected is determined. A schematic diagram of the gearbox is displayed (based on the structural diagram display unit in the gearbox visualization inspection module: according to the selected gearbox model in the object to be inspected determination results, a schematic diagram of the internal structure is dynamically loaded and displayed, with clickable inspection locations marked on the diagram). Linear layout conditions are also provided (linear layout: the marking points are distributed along the axial direction of the gear shaft, forming a one-dimensional linear relationship (e.g., bearings A and B are installed sequentially along the axial direction, and the marking points are on the same axis). The correlation of axial marking points strongly depends on axial alignment (e.g., the fit clearance between the bearing inner ring and the shaft, the axial spacing between bearings). Positional deviations of these marking points (e.g., incorrect axial positioning) directly lead to axial dimensional chain errors (e.g., incorrect calculation of bearing axial clearance). (Incorrect) For example, if there is a positioning deviation in the axial marking points on the inner ring of the bearing, it will misjudge the axial clearance of the bearing (too large or too small). Bearing clearance is a core cause of excessive vibration and accelerated wear, which may lead maintenance personnel to mistakenly judge "bearing damage" or "no maintenance required," thus causing failure. (Plane layout conditions: marking points are distributed in the same plane, forming a two-dimensional planar relationship (such as marking points on the meshing tooth surface of the gear end face, or multiple marking points on the circumference of the bearing outer ring). The correlation of planar marking points strongly depends on flatness and circumferential uniformity (such as the contact spots on the gear tooth surface requiring uniform distribution in the same plane, and the marking points on the bearing outer ring requiring circumferential symmetry). Such positional deviations or incorrect relationships of marking points will directly lead to Failure of planar geometric constraints (such as the contact area of ​​gear meshing tooth surfaces, the fit between the bearing outer ring and the housing), for example, uneven distribution of planar marking points on the bearing outer ring can lead to positioning deviations of vibration sensor mounting points. The collected vibration signals cannot accurately reflect the circumferential load distribution of the bearing outer ring, thus misjudging "unilateral bearing wear," when in fact it is signal distortion caused by sensor position deviation. Spatial layout relationships (three-dimensional spatial layout: marking points are distributed in three-dimensional space, satisfying spatial geometric constraints (such as the coaxiality of the gear shaft and bearing, the center distance and parallelism of gear meshing, the spatial orientation of lubrication lines and the spatial positional relationship between the gear or bearing) are the most complex spatial relationships in a gearbox and the core correlation that fully reflects the operating status of the components.The correlation between 3D spatial annotation points strongly depends on the comprehensive constraints of spatial position (such as coaxiality, parallelism, and spatial distance tolerance). Positional deviations or incorrect relationships of these annotation points can lead to the coupled propagation of 3D geometric errors—errors in a single dimension (such as radial deviation) can superimpose with axial errors, affecting rotational accuracy and power transmission. For example, in a three-stage gear pair of a large wind turbine gearbox, the annotation points must simultaneously meet the requirements of axial spacing (parallelism), radial center distance (coaxiality), and vertical meshing clearance. If the 3D position of one annotation point is misjudged during positioning, it can cause the center distance or parallelism to exceed tolerances, leading to abnormal gear meshing noise. However, the system may misjudge this as a "bearing failure" rather than an installation error in the gear pair, resulting in incorrect maintenance and delayed fault handling. Axial deviation, flatness deviation, and coaxial deviation can also be addressed (e.g., by recording the actual positional deviation of the annotation points using high-precision positioning equipment such as laser positioning instruments or 3D scanners, and having the user upload the detected deviation data to the gearbox visualization inspection module for subsequent analysis of misjudgment rates).

[0057] The specific step S2 includes the following sub-steps:

[0058] Collect environmental temperature, humidity, wind speed, and vibration data of the gearbox in the test object determination results to obtain the characteristic data of the wind farm to be tested.

[0059] Historical monitoring data is obtained, and the correlation influence coefficient between historical deviation values ​​and historical wind farm characteristic data is obtained based on the correlation influence trend characteristics between historical deviation values ​​and historical wind farm characteristic data.

[0060] Based on the characteristic data of the wind farm to be measured and the correlation influence coefficient, the pre-compensation deviation value is obtained.

[0061] The actual deviation value is obtained by subtracting the measured deviation value from the pre-compensated deviation value.

[0062] A misjudgment prediction model is constructed based on historical deviation values, historical misjudgment rates in historical detection data, and historical gearbox operation data in historical detection data.

[0063] The operating parameters of the gearbox in the determination results of the object to be inspected are collected to obtain the inspection data to be tested. The inspection data to be tested and the actual deviation value are input into the misjudgment prediction model for testing to obtain the current misjudgment rate.

[0064] A preset false positive threshold is set. If the current false positive rate is less than the false positive threshold, there is no need to expand the initial annotation checkpoints. Based on the initial annotation checkpoints, the on-site images uploaded by the user are collected to obtain the image information to be tested.

[0065] If the current misjudgment rate is greater than or equal to the misjudgment threshold, the initial annotation checkpoint is extended and expanded according to the straight line layout position condition to obtain amplified annotation point one; or, according to the planar layout position condition, the initial annotation checkpoint is expanded diagonally to obtain amplified annotation point two; or, according to the spatial layout position condition, the initial annotation checkpoint is expanded layer by layer to obtain amplified annotation point three. The amplified annotation point one, amplified annotation point two, and amplified annotation point three are combined to form the amplified annotation checkpoint.

[0066] Obtain the additional checkpoints selected by the user from the augmented annotation checkpoints. Based on the additional checkpoints, collect the on-site images uploaded by the user to obtain the second image information to be tested.

[0067] The image information to be tested, part one, and the image information to be tested, part two, are combined to form preprocessed image information.

[0068] Specifically, the data includes characteristic data of the wind farm under test (based on the wind farm and gearbox management module, uploaded by the user, including ambient temperature and humidity (detected by temperature and humidity sensors such as wall-mounted temperature and humidity transmitters and portable temperature and humidity detectors), wind speed (detected by wind speed sensors such as cup anemometers and ultrasonic anemometers), and vibration data (detected by vibration sensors such as piezoelectric vibration sensors and accelerometers)), and correlation influence coefficients (e.g., Y=a*k1+b*k2+c*k3+d**k4+U, where Y refers to the historical deviation value and a refers to the temperature data in the historical wind farm characteristic data). k1 refers to the correlation coefficient between historical deviation values ​​and temperature data; b refers to humidity data in historical wind farm characteristic data; k2 refers to the correlation coefficient between historical deviation values ​​and humidity data; c refers to wind speed data in historical wind farm characteristic data; k3 refers to the correlation coefficient between historical deviation values ​​and wind speed data; d refers to vibration signal data in historical wind farm characteristic data; k4 refers to the correlation coefficient between historical deviation values ​​and vibration signal data; and U refers to the additional error. Substituting the known historical deviation values ​​and historical wind farm characteristic data yields the following result. The pre-compensation deviation value (e.g., by substituting the wind farm characteristic data and the correlation influence coefficient into Y=a*k1+b*k2+c*k3+d**k4+U, the pre-compensation deviation value can be obtained), constructing a misjudgment prediction model (e.g., by matching historical deviation values, historical misjudgment rates, and historical gearbox operation data into a data matching table G: historical deviation value - historical misjudgment rate - historical gearbox operation data), current misjudgment rate (by comparing the test data and actual deviation values ​​with the data matching table G using known data comparison, the corresponding misjudgment rate is obtained), and preset... False positive threshold (set manually based on historical data and updated in real time), image information to be tested (based on the photo integration management unit in the gearbox visualization inspection module: provides one-click screenshot and photo upload functions, automatically associates the captured photos with the initial marked checkpoints, and displays the number of photos in real time on the button at that location), expanded marked checkpoints (if the current false positive rate is greater than or equal to the false positive threshold, it means that the deviation in this case has greatly affected the judgment result of subsequent inspections, and data correction is required, i.e., reselecting the marked checkpoint locations and performing a second inspection).Expanding Annotation Point 1: Extend and expand the existing annotation points along the straight lines. In the axial range where the deviation value exceeds the standard, densify the annotation points (e.g., reduce the original 5cm interval to 2cm), focusing on expanding the area where the deviation peak is located and its surroundings, accurately capturing the specific distribution pattern of the axial deviation; Expanding Annotation Point 2: For the grid area where the flatness deviation exceeds the standard, add diagonal annotation points to form a composite layout of "matrix + diagonal", accurately locating the specific position of the planar concave and convex deformation; Expanding Annotation Point 3: Based on the annotation points laid out in the three-dimensional dot matrix, add layered annotation points around the outline of the component with excessive deviation (e.g., add layers along the component height direction), to achieve multi-dimensional accurate capture of three-dimensional spatial deviation; Image Information 2 (based on Image Information 1, and so on).

[0069] The specific step S3 includes the following sub-steps:

[0070] Based on the amplified and marked checkpoints, the operating parameters of the gearbox in the determination results of the object to be inspected are collected to obtain additional inspection data. The additional inspection data and the inspection data to be tested are combined to form preprocessed inspection data.

[0071] Based on historical detection data, a pre-set abnormal keyword rule base is used to identify anomalies in the pre-processed inspection data and obtain the identification results. If the identification result is abnormal and matches the abnormal keyword rule base, an abnormal record is created and the abnormal record information is output.

[0072] The preprocessed image information, preprocessed inspection data, and abnormal record information are stored, synchronized, and their consistency is verified to obtain the inspection data archive.

[0073] The system obtains the object and time range selected by the user for generating the report, selects relevant data from the inspection data archive based on the selected object and time range, generates a standardized professional inspection report, sends the professional inspection report to the user, and outputs the full-process inspection results.

[0074] Specifically, additional inspection data (based on the inspection data to be tested, and so on) and a pre-set abnormal keyword rule base (based on the rule engine unit in the abnormal intelligent identification and management module: pre-set abnormal categories (such as bearing abnormalities, gear wear) and associated keyword rule bases; the built-in rule engine transforms expert experience into executable judgment rules: for example, bearing abnormality category: the pre-set abnormal keyword rule base includes keywords such as "bearing," "roller," "raceway," "scratches," "pitting," "abnormal noise," and "jamming," and the judgment rule is set as "if the inspection description contains 'bearing' + any one of the fault keywords, it is judged as a bearing abnormality"; gear wear category: the pre-set keyword rule base includes keywords such as "gear," "meshing," "wear," "tooth surface," "scratches," and "tooth tip wear," and the judgment rule is set as "if the inspection description contains 'gear' + any one of the fault keywords, it is judged as a bearing abnormality"; Any wear-related keyword is used to determine abnormal gear wear. When a user enters an inspection description (such as "radial scratches on the bearing rollers, slight abnormal noise during operation") in the intelligent inspection form unit of the gearbox visualization inspection module, the rule engine unit captures the keywords "bearing," "roller," "scratches," and "abnormal noise" in the description in real time, matches them with the keyword rule library corresponding to bearing abnormalities, triggers the abnormality judgment rule, automatically identifies the potential abnormality corresponding to the inspection description as "bearing abnormality," and feeds the identification result back to the automatic identification unit. The automatic identification unit creates an abnormality record of the "bearing abnormality" category, completing the automatic identification and classification of abnormalities and realizing the standardized and automated application of expert experience. The intelligent inspection form unit in the gearbox visualization inspection module includes a status assessment (such as: no obvious abnormality, observe carefully, need to be replaced), an inspection description text box, and an operation and maintenance suggestion input area. The inspection description text box is a "smart text input box" that can display standardized damage types (such as "roller radial scratches" and "raceway pitting") in real time based on user-input keywords (such as "bearing") for selection. This ensures standardized terminology and guarantees consistency and accurate expression in all user-entered inspection descriptions, preventing data confusion caused by differences in descriptions from different inspectors (such as mixing up "scratches" and "scratches"). When generating subsequent report information, the "inspection description" is extracted from the unified data management module. The data, precisely the content standardized by this function—standardized inspection descriptions can be directly filled into the "Inspection Result Table" and "Maintenance Suggestions" sections of the report, ensuring the report content is standardized and professional, while avoiding data corruption due to inconsistent descriptions. Abnormal record information (based on the automatic identification unit in the abnormality intelligent identification and management module: real-time monitoring of inspection data (preprocessing inspection data); when the inspection status is abnormal and the description text matches the rules, an abnormal record is automatically created and initially categorized); based on the process management unit in the abnormality intelligent identification and management module: full-process tracking management of automatically or manually created abnormal record information, including review, allocation, processing, and closed-loop management.The inspection data archive (based on the unified data management module in the anomaly intelligent identification and management module: serving as the data hub of the wind power equipment inspection platform, providing access, query, and synchronization services for wind farm data, gearbox data (pre-processed inspection data), inspection result data (anomaly results), photo index data (pre-processed image information), and anomaly record data to all business modules through a unified interface, ensuring data consistency and forming a complete inspection data archive), generates standardized professional inspection reports (based on the report generation unit in the intelligent report generation and email sending module: providing a report parameter interface (selecting wind farm, gearbox, time, inspector, etc.), based on preset Wo... The report template automatically extracts relevant data from the data management module and populates it to generate a complete inspection report document including a cover, preface, inspection results table, maintenance suggestions, and embedded photos. Users can click to enter the report generation unit of the intelligent report generation and email sending module. This unit provides a visual report parameter interface (based on the email sending unit in the intelligent report generation and email sending module). Users can complete the parameter selection in the interface in sequence: select the wind farm to be reported (e.g., "XX Wind Farm"); select the gearbox to be inspected under the wind farm (e.g., "No. 1 Wind Turbine Gearbox"); and set the report time range (e.g., "March 1, 2026 - March 31, 2026"). Fill in the inspector information (e.g., "Zhang San"); after parameter selection, the system automatically triggers the data extraction process, extracting all relevant data within the parameter range from the unified data management module, including: basic information of the wind farm and gearbox (manufacturer, model, etc.), all inspection results within the time period (status assessment, description, etc. of each marked inspection point), related photos uploaded on-site, anomaly records, and maintenance suggestions. Then, all extracted data is automatically filled into the corresponding positions according to the format requirements of the preset Word report template, ultimately generating a complete standardized professional inspection report. The report cover is automatically filled with information such as wind farm, gearbox, time, and inspector; the preface summarizes the scope and purpose of this inspection; the inspection results table summarizes the inspection data of each inspection point; maintenance suggestions provide targeted opinions for corresponding anomaly records; and on-site photos are embedded into the corresponding chapters of the report in the order of inspection location, forming a complete and directly usable inspection report document (i.e., a professional inspection report). Email sending unit: provides a visual interface, allowing users to configure the email server, sender and recipient, subject, and body, and can send the generated report file and other local files as attachments.

[0075] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A method for inspection and maintenance management of wind power generation equipment, characterized in that, Includes the following steps: Step S1: Obtain the basic information of the wind turbine generator set and corresponding gearbox selected by the user, output the result of the object to be inspected, display the gearbox schematic diagram according to the result of the object to be inspected, and collect the axial deviation, flatness deviation and coaxiality deviation according to the positional relationship of the initial marked inspection point selected by the user on the gearbox schematic diagram, which belongs to the straight line layout, plane layout and three-dimensional space layout, and obtain the measured deviation value. Step S2: Based on the correlation and trend characteristics between historical deviation values ​​and historical wind farm characteristic data in historical detection data, error compensation is performed on the deviation value to be measured to obtain the actual deviation value. Based on historical detection data, a misjudgment prediction model is established. The actual deviation value is input into the misjudgment prediction model for testing to obtain the current misjudgment rate. Based on the current misjudgment rate, the initial marked checkpoints are expanded to obtain expanded marked checkpoints. Based on the initial marked checkpoints and expanded marked checkpoints, on-site images uploaded by users are collected, and preprocessed image information is obtained by combining them. Step S3: Based on the initial and expanded checkpoints, collect the operating parameters of the gearbox in the determination results of the object to be inspected, and obtain the preprocessed inspection data. Based on the preprocessed image information and the preprocessed inspection data, obtain the inspection data archive. Based on the inspection data archive and the object and time range selected by the user to generate the report, output the full-process inspection results.

2. The inspection method for operation and maintenance management of wind power generation equipment according to claim 1, characterized in that, Step S1 includes: Obtain the user's identity authentication information when entering the wind power equipment inspection platform; obtain the user's permission verification information based on the identity authentication information; display the main interface of the wind power equipment inspection platform to the user based on the permission verification information; and output the display results of the main interface of the platform. The user selects the basic information of the wind turbine generator set to be inspected and the corresponding gearbox from the results displayed on the main interface of the platform, and the result of the inspection object determination is output. Based on the gearbox model selected from the results of the test object determination, display the internal structure diagram of the gearbox and output the gearbox schematic diagram.

3. The inspection method for operation and maintenance management of wind power generation equipment according to claim 2, characterized in that, Step S1 also includes: Get the initial annotation checkpoints selected by the user on the gearbox schematic diagram. If all the initial annotation checkpoints are arranged in a straight line, output the straight line arrangement position condition. Based on the straight layout position conditions, the axial deviation of the gearbox in the determination result of the object to be inspected is collected to obtain the axial deviation value; If all checkpoints in the initial checkpoints are arranged in a planar layout, output the planar layout position conditions. Based on the planar layout conditions, the flatness deviation of the gearbox in the determination result of the object to be inspected is collected to obtain the flatness deviation value; If all the initial annotation checkpoints are arranged in a three-dimensional spatial relationship, output the spatial arrangement position conditions. Based on the spatial layout conditions, the coaxiality deviation of the gearbox in the determination result of the object to be inspected is collected to obtain the three-dimensional spatial deviation value. The axial deviation value, flatness deviation value, and three-dimensional spatial deviation value are combined to form the deviation value to be measured.

4. The inspection method for operation and maintenance management of wind power generation equipment according to claim 3, characterized in that, Step S2 includes: Collect environmental temperature, humidity, wind speed, and vibration data of the gearbox in the test object determination results to obtain the characteristic data of the wind farm to be tested. Historical detection data is obtained, and the correlation influence coefficient between historical deviation values ​​and historical wind farm characteristic data is obtained based on the correlation influence trend characteristics between the historical deviation values ​​and historical wind farm characteristic data in the historical detection data.

5. The inspection method for operation and maintenance management of wind power generation equipment according to claim 4, characterized in that, Step S2 also includes: Based on the characteristic data of the wind farm to be tested and the correlation influence coefficient, the pre-compensation deviation value is obtained; The actual deviation value is obtained by subtracting the measured deviation value from the pre-compensation deviation value. Based on the historical deviation values, the historical misjudgment rate in the historical detection data, and the historical gearbox operation data in the historical detection data, a misjudgment prediction model is constructed. The operating parameters of the gearbox in the determination result of the object to be inspected are collected to obtain the inspection data to be tested. The inspection data to be tested and the actual deviation value are input into the misjudgment prediction model for testing to obtain the current misjudgment rate.

6. The inspection method for operation and maintenance management of wind power generation equipment according to claim 5, characterized in that, Step S2 also includes: A preset false judgment threshold is set. If the current false judgment rate is less than the false judgment threshold, there is no need to expand the initial annotation checkpoints. Based on the initial annotation checkpoints, the on-site images uploaded by the user are collected to obtain the image information to be tested. If the current misjudgment rate is greater than or equal to the misjudgment threshold, the initial annotation checkpoint is extended and expanded according to the linear layout position condition to obtain amplified annotation point one; or, the initial annotation checkpoint is expanded diagonally according to the planar layout position condition to obtain amplified annotation point two; or, the initial annotation checkpoint is expanded layer by layer according to the spatial layout position condition to obtain amplified annotation point three. The amplified annotation point one, amplified annotation point two, and amplified annotation point three are combined to form the amplified annotation checkpoint. Obtain additional checkpoints selected by the user from the amplified annotation checkpoints, and collect on-site images uploaded by the user based on the additional checkpoints to obtain the second image information to be tested; The image information to be tested, part one, and the image information to be tested, part two, are combined to form preprocessed image information.

7. The inspection method for operation and maintenance management of wind power generation equipment according to claim 6, characterized in that, Step S3 includes: Based on the amplified and marked checkpoints, the operating parameters of the gearbox in the determination results of the object to be inspected are collected to obtain additional inspection data. The additional inspection data and the inspection data to be tested are combined to form preprocessed inspection data. Based on historical detection data, a pre-set abnormal keyword rule base is used to identify anomalies in the pre-processed inspection data and obtain identification results. If the identification result is abnormal and hits the abnormal keyword rule base, an abnormal record is created and the abnormal record information is output. The preprocessed image information, preprocessed inspection data, and abnormal record information are stored, synchronized, and their consistency is verified to obtain the inspection data archive. The system obtains the object and time range selected by the user for generating the report, selects relevant data from the inspection data archive based on the selected object and time range, generates a standardized professional inspection report, transmits the professional inspection report to the user, and outputs the full-process inspection results.

Images