Fan main shaft full life cycle tracing method and system based on digital twinning

Abstract

This invention discloses a method and system for tracing the entire lifecycle of a wind turbine main shaft based on digital twins, relating to the field of data management technology. The method involves collecting physical data of the wind turbine main shaft and constructing a digital twin model of the main shaft based on the physical data and performance model. The entire lifecycle data of the wind turbine main shaft is collected and stored in corresponding stage channels to obtain stage data. Multiple root points are set in the stage channels, and the stage data is bound based on these root points. When an anomaly occurs in a stage channel, the data in the root points is transferred. The stage data is then constructed into the digital twin model for simulation based on the stage channels. This invention ensures data security even during long-term storage and allows for direct use in fault analysis and tracing of the wind turbine main shaft, avoiding incorrect fault tracing due to insecure stored data.

Description

Technical Field

[0001] This invention relates to the field of data management technology, specifically to a method and system for tracing the entire lifecycle of wind turbine main shafts based on digital twins. Background Technology

[0002] Currently, the world is vigorously promoting the transformation of its energy structure towards clean energy. Wind power, as a new energy source with strong stability and high exploitability, is experiencing rapid and continuous expansion in installed capacity. However, wind turbines are mostly deployed in complex and harsh environments such as the field and offshore. As the core component of the wind turbine's drivetrain, the main shaft's operational stability directly determines the wind power generation efficiency and safety. Therefore, it is necessary to remotely monitor it and collect operational data to analyze the working status and fault conditions of the wind turbine main shaft. Currently, data collection for the wind turbine main shaft throughout its lifecycle is a long-term process. During this process, a large amount of data needs to be stored. Data storage is susceptible to interference from external networks, which can easily lead to inaccuracies and consequently, inaccurate tracing of wind turbine main shaft faults. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for tracing the entire lifecycle of a wind turbine main shaft based on digital twins, so as to solve the problems in the background technology.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a method for tracing the entire lifecycle of a wind turbine main shaft based on digital twins, comprising the following steps:

[0005] Collect physical data of the wind turbine main shaft, and build a digital twin model of the wind turbine main shaft based on the physical data and performance model;

[0006] The entire lifecycle data of the wind turbine main shaft is collected and stored in the corresponding stage channel to obtain stage data. Multiple root points are set in the stage channel, and the stage data is bound based on the root points.

[0007] When an anomaly occurs in the stage channel, the data in the root point is transferred and the transferred data is associated with the tactile paving. After the stage channel is normal, the stage data in the stage channel is restored through the root point.

[0008] Based on the stage channel, stage data is built into a digital twin model for simulation. When the main shaft of the wind turbine fails, key data at the time of the failure is extracted. Based on the stage channel, the relevant data corresponding to the key data is traced, and the relevant data is analyzed to obtain the root cause data of the failure.

[0009] In a preferred embodiment, the step of collecting physical data of the wind turbine main shaft and building a digital twin model of the wind turbine main shaft based on the physical data and performance model includes:

[0010] The basic model is constructed based on physical data, which includes geometric data, material mechanical parameters, and thermodynamic properties.

[0011] The performance model is fused with the basic model to obtain a digital twin model of the wind turbine main shaft, which is then stored in the traceability platform.

[0012] In a preferred embodiment, the step of collecting and storing the full lifecycle data of the wind turbine main shaft into the corresponding stage channel to obtain stage data, setting multiple root points in the stage channel, and binding the stage data based on the root points includes:

[0013] Multiple data collection points are set up for the main shaft of the wind turbine. Based on the data collection points, corresponding receiving ports are set up in the digital twin model, and a communication connection is established between the data collection points and the corresponding receiving ports.

[0014] In the traceability platform, corresponding stage channels are set up for the entire life cycle stage of the wind turbine main shaft. The operating data of the wind turbine main shaft is collected based on the collection points and marked with timestamps. The operating data marked with timestamps is transmitted to the corresponding receiving port and stored in the stage channel of the traceability platform as stage data until the stage data corresponding to the entire life cycle stage is obtained, which is used as the entire life cycle data of the wind turbine main shaft.

[0015] Multiple root points are set in the stage channel, and stage data in the stage channel is captured and stored through the root points.

[0016] In a preferred embodiment, the step of setting multiple root points in the stage channel and capturing and storing stage data in the stage channel through the root points includes:

[0017] In the phased passageway, a main passageway and a tactile paving are set up, and the tactile paving and the main passageway are connected.

[0018] Multiple root points are set in the main channel. The root points include internal points, external points and switching ports. The switching ports are set in the main channel. The internal points and external points are connected through the switching ports. Each internal point includes a moving point. The moving point corresponds to the switching port connected to its internal point. When there is no abnormality in the stage channel, the switching port is in the closed state.

[0019] A correlation diagram is set up in the tactile paving. The correlation diagram consists of multiple correlation points. Each correlation point is connected to a movement point in a one-to-one correspondence. When the inner point where the movement point is located does not store stage data, the correlation point and the movement point are disconnected.

[0020] When storing stage data in a stage channel, the stage data is classified based on the data type to obtain multiple stage type data. The storage range of the main channel in the corresponding stage channel is divided according to the data type, and multiple storage type ranges of the corresponding data type are obtained respectively.

[0021] Based on the timestamp, the stage type data is stored in the corresponding storage type range of the preset time period, thus completing the capture and storage of the stage data.

[0022] After storing the stage type data of the preset time period in the inlier, the connection relationship between the moving point in the inlier and the corresponding associated point in the association graph is enabled, and the associated points in the association graph are divided and associated according to the inlier corresponding to the storage type range.

[0023] Fill the gap between the main channel and the interior point with invalid data.

[0024] In a preferred embodiment, the step of transferring data from the root point when an anomaly occurs in the stage channel, associating the transferred data with the tactile paving, and restoring the stage data in the stage channel through the root point after the stage channel returns to normal includes:

[0025] The main channel of the corresponding stage channel is configured with an access port, and the corresponding access port is marked with an authorized network. When an unauthorized network accesses the main channel, it indicates that there is an anomaly in the stage channel.

[0026] When an anomaly occurs in the stage channel, the switching port is opened to transfer the movement point in the inner point and the stage type data of the preset time period to the outer point through the switching port, and the switching port is closed to disconnect the outer point from the main channel, while always maintaining the connection between the movement point and the corresponding associated point.

[0027] When an unauthorized network exits access, it indicates that the phase channel is normal. The external point is reconnected to the corresponding switching port based on the mobile point, and the mobile point and the phase type data of the preset time period in the external point are transferred to the internal point through the switching port to complete the recovery of the phase data in the main channel.

[0028] In a preferred embodiment, the step of constructing stage data into a digital twin model based on stage channels for simulation, extracting key data at the time of the fault when the wind turbine main shaft fails, tracing relevant data corresponding to the key data based on stage channels, and analyzing the relevant data to obtain the root cause data of the fault includes:

[0029] In the digital twin model, the receiving port receives stage data and stores it in the corresponding stage channel;

[0030] The simulation is driven by the digital twin model based on the operating data of the wind turbine main shaft. When the wind turbine main shaft fails, the key data at the time of the failure is extracted, and the relevant data corresponding to the key data is traced based on the stage channel.

[0031] Fault correlation analysis was performed on the relevant data to obtain the root cause data of the faults.

[0032] This invention also provides a digital twin-based full lifecycle traceability system for wind turbine main shafts, including:

[0033] The module is used to collect physical data of the wind turbine main shaft and build a digital twin model of the wind turbine main shaft based on the physical data and performance model;

[0034] The configuration module, connected to the construction module, is used to collect and store the full life cycle data of the wind turbine main shaft into the corresponding stage channel to obtain stage data. Multiple root points are set in the stage channel, and the stage data is bound based on the root points.

[0035] The transfer module, connected to the setting module, is used to transfer the data in the root point when there is an anomaly in the stage channel, and associate the transferred data with the tactile paving. After the stage channel is normal, the stage data in the stage channel is restored through the root point.

[0036] The analysis module, connected to the setting module, is used to build stage data into a digital twin model for simulation based on stage channels. When a fault occurs in the main shaft of the wind turbine, the key data at the time of the fault is extracted, and the relevant data corresponding to the key data is traced based on the stage channels. The relevant data is analyzed to obtain the root cause data of the fault.

[0037] In a preferred embodiment, the building module includes:

[0038] The data acquisition unit is used to construct the basic model based on physical data, which includes geometric data, material mechanical parameters, and thermodynamic properties.

[0039] The fusion unit is used to merge the performance model with the basic model to obtain a digital twin model of the wind turbine main shaft and store it in the traceability platform.

[0040] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0041] This invention provides better protection for stage data at the root point through stage channels, preventing data from being obtained or tampered with by unauthorized networks. Data from the entire lifecycle of the wind turbine main shaft needs to be stored for a long time, ensuring that the data remains secure even after long-term storage. This data can be directly used for fault analysis and fault tracing of the wind turbine main shaft, avoiding incorrect fault tracing of the wind turbine main shaft due to insecure stored data. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0043] Figure 1 This is a flowchart of the method of the present invention.

[0044] Figure 2 This is a system block diagram of the present invention. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] Example 1, please refer to Figure 1 As shown in this embodiment, the wind turbine main shaft full lifecycle traceability method based on digital twins includes the following steps:

[0047] S1. Collect physical data of the wind turbine main shaft and build a digital twin model of the wind turbine main shaft based on the physical data and performance model;

[0048] S2. Collect and store the full life cycle data of the wind turbine main shaft into the corresponding stage channel to obtain stage data. Set multiple root points in the stage channel and bind the stage data based on the root points.

[0049] S3. When there is an anomaly in the stage channel, the data in the root point is transferred and the transferred data is associated with the tactile paving. After the stage channel is normal, the stage data in the stage channel is restored through the root point.

[0050] S4. Based on the stage channel, the stage data is built into the digital twin model for simulation. When the main shaft of the wind turbine fails, the key data at the time of the failure is extracted. Based on the stage channel, the relevant data corresponding to the key data is traced, and the relevant data is analyzed to obtain the root cause data of the failure.

[0051] As described in steps S1-S4 above, the stage data in the root point can be better protected through the stage channel, preventing the data from being obtained or tampered with by unauthorized networks. The data of the entire life cycle of the wind turbine main shaft needs to be stored for a long time, so that the data can still be secure under long-term storage and can be directly used for fault analysis and fault tracing of the wind turbine main shaft, avoiding incorrect fault tracing of the wind turbine main shaft due to the security of stored data.

[0052] In one embodiment, step S1, which involves collecting physical data of the wind turbine main shaft and building a digital twin model of the wind turbine main shaft based on the physical data and performance model, includes:

[0053] S11. Construct a basic model based on physical data, which includes geometric data, material mechanical parameters, and thermodynamic properties.

[0054] S12. Integrate the performance model with the basic model to obtain a digital twin model of the wind turbine main shaft and store it in the traceability platform;

[0055] As described in steps S11-S12 above, when establishing a digital twin model, it is necessary to clarify the scope and priority of data collection, prioritizing the acquisition of three core data categories: design, materials, and processing, and then supplementing with auxiliary information such as suppliers and testing institutions. Design phase data: Extract structural dimensions, strength verification parameters, and fatigue life simulation values ​​from CAD design documents and simulation reports, while simultaneously recording design versions and review comments. Material composition data: Collect batch reports and composition testing data of spindle raw materials, such as carbon and chromium content; mechanical performance parameters, tensile strength, and hardness, and bind them to material batch numbers. Processing technology data: Extract processing equipment parameters from the production system, such as cutting speed and feed rate; heat treatment process assembly steps, such as temperature and holding time; and associate processing stations with time nodes. Data verification and completion: Verify the accuracy of data by comparing it with design standards and industry specifications. Complete missing data by tracing raw material suppliers and production logs. Specific required data can be collected based on actual needs. Data classification and structuring: Classify the collected data into geometric, physical, and performance categories, enter them into the database, and establish associations, such as binding geometric dimensions with processing parameters. Geometric Model Construction: Based on CAD design data, a 3D geometric model is built using UG to recreate the spindle structure, mating relationships, and key dimensions. Physical Model Integration: Material mechanical parameters and thermodynamic properties are imported, and combined with finite element analysis tools such as ANSYS to construct a physical model that simulates the spindle's stress and temperature conduction characteristics. Performance Model Fusion: Design simulation data and historical operating performance data are integrated to establish a performance degradation model, linking load, rotational speed, and lifespan loss. Model Calibration: Model parameters are adjusted using actual physical spindle testing data to ensure accurate mapping between the digital twin and the physical spindle, such as factory inspection dimensions and no-load operating vibration values. Data Transmission Requirements: The types of operating data to be received in real-time are determined, including vibration, temperature, and load; the transmission frequency and data format (JSON, XML), such as 1 time / second. Data field naming rules, unit standards, and abnormal data handling methods are defined, such as missing value filling and outlier removal, to ensure data consistency; finally, the constructed digital twin model is obtained and stored in the traceability platform.

[0056] In one embodiment, step S2, which involves collecting and storing the full lifecycle data of the wind turbine main shaft into the corresponding stage channel to obtain stage data, setting multiple root points in the stage channel, and binding the stage data based on the root points, includes:

[0057] S21. Set multiple data collection points for the main shaft of the wind turbine, set corresponding receiving ports in the digital twin model based on the data collection points, and establish communication connections between the data collection points and the corresponding receiving ports.

[0058] S22. In the traceability platform, set up corresponding stage channels for the full life cycle stages of the wind turbine main shaft, collect the operating data of the wind turbine main shaft based on the collection points and mark the timestamp, transmit the operating data marked with timestamps to the corresponding receiving port, and store it in the stage channel of the traceability platform as stage data until the stage data corresponding to the full life cycle stages is obtained, which is used as the full life cycle data of the wind turbine main shaft.

[0059] S23. Set multiple root points in the stage channel, and use the root points to capture and store the stage data in the stage channel.

[0060] In one embodiment, step S23, which involves setting multiple root points in a stage channel and capturing and storing stage data in the stage channel through these root points, includes:

[0061] S231. A main passage and a tactile paving are set up in the stage passage, and the tactile paving and the main passage are connected.

[0062] S232. Multiple root points are set in the main channel. The root points include internal points, external points and switching ports. The switching ports are set in the main channel. The internal points and external points are connected through the switching ports. Each internal point includes a mobile point. The mobile point corresponds to the switching port connected to the internal point (there is a binding relationship between the mobile point and the switching port connected to the internal point). When there is no abnormality in the stage channel, the switching port is in the closed state.

[0063] S233. Set up an association diagram in the tactile paving. The association diagram consists of multiple association points. Each association point is connected to a movement point in a one-to-one correspondence. When the inner point where the movement point is located does not store stage data, the association point and the movement point are disconnected.

[0064] S234. When storing stage data in a stage channel, the stage data is classified based on the data type to obtain multiple stage type data. The storage range of the main channel in the corresponding stage channel is divided according to the data type to obtain multiple storage type ranges of the corresponding data type.

[0065] S235. Based on the timestamp, store the stage type data in the corresponding storage type range of the preset time period in the inner point of the corresponding storage type range to complete the capture and storage of the stage data.

[0066] S236. After storing the stage type data of the preset time period in the inner point, enable the connection relationship between the moving point in the inner point and the corresponding associated point in the association diagram, and divide the associated points in the association diagram into associations according to the inner points corresponding to the storage type range.

[0067] S237. Fill invalid data between the main channel and the inner point.

[0068] As described in steps S21-S23 above, after the fan main shaft is installed and used, in order to detect the status of the fan main shaft, multiple acquisition points need to be set for the fan main shaft. These acquisition points include devices such as a main shaft running vibration acquisition device, a temperature sensor, and a load acquisition device. Multiple acquisition points are used to collect corresponding data for subsequent status analysis and fault tracing of the fan main shaft. In the traceability platform, corresponding stage channels are set up for the entire life cycle of the wind turbine main shaft. For example, the entire life cycle of the wind turbine main shaft includes the design stage, manufacturing stage, operation and maintenance stage, and scrapping stage. Based on these stages, corresponding stage channels are set up in the traceability platform, where the stage channels are storage spaces within the traceability platform. Then, the operating data of the wind turbine main shaft is collected and timestamped through collection points. For example, the temperature sensor collects the temperature of the wind turbine main shaft and marks the collection time (timestamp). The operating data marked with the timestamp is transmitted to the corresponding receiving port. The location of the receiving port in the digital twin model is the same as the location of the wind turbine main shaft operating data collected by the collection point. For example, the temperature sensor collects the data at the mating part of the main shaft and the bearing. Since the bearing is the core rotating component of the main shaft, it is prone to overheating due to insufficient lubrication, wear, and installation deviation. Abnormal temperature directly reflects bearing failure, such as jamming or burning, which affects the life of the main shaft. Thus, the temperature data collected by the temperature sensor is transmitted to the receiving port at the corresponding location in the digital twin model. The location of the receiving port in the digital twin model is also set according to the collection point's location on the actual wind turbine main shaft. The collected, timestamped operational data is then stored in the stage channels of the traceability platform as stage data, until the stage data corresponding to the entire lifecycle stage is obtained, serving as the full lifecycle data for the wind turbine main shaft. Multiple root points are then set in each stage channel to capture and store the stage data. A main channel and a blind channel are set within each stage channel, connected to the main channel. The blind channel is a storage space within the traceability platform independent of the stage channels. When not accessed by an unauthorized network, the blind channel is connected to the main channel; conversely, it is disconnected to prevent unauthorized network access.The main channel contains multiple root points, each composed of an inner point, an outer point, and a switch port. Within the inner points, there are also moving points. All inner, outer, and moving points are virtual machines. Outer points are virtual machines located in the external storage space of the main channel, while inner points are virtual machines located in the internal storage space of the main channel. They are connected via switch ports. An association diagram is set up in the tactile paving, consisting of multiple association points, each a virtual machine. It should be noted that the settings for association points and inner points are based on selecting storage addresses in the storage space. Association points are set in the tactile paving, and the storage locations of the switch ports corresponding to the inner points are configured on the main channel. The movement points are connected one-to-one, allowing the relationships between root points in the main channel to be directly represented through the association diagram. When the internal point containing the movement point does not store stage data, the association point and the movement point are disconnected. When stage data is stored in the stage channel, the stage data is classified based on data type, resulting in multiple stage type data. For example, in the operation and maintenance stage, the collected operational data includes operational vibration, temperature, and load data. Therefore, the operation and maintenance stage data needs to be classified according to this data type. The storage range of the main channel in the corresponding stage channel is divided according to the data type, resulting in multiple storage type ranges for each corresponding data type. Each storage type range contains multiple root points. Based on timestamps, stage-type data is stored in inner points within the corresponding storage type range according to preset time periods. This chronological storage within inner points is then linked and sorted to complete the capture and storage of stage data. This capture and storage refers to the inner points capturing and storing stage data, not storing the entire stage data in the main channel. After the preset time period of stage-type data is stored in the inner points, the connection between the movement points in the inner points and the corresponding associated points in the association diagram is enabled. The associated points in the association diagram are divided and linked according to the corresponding inner points within the storage type range. The range in the tactile paving also corresponds to the set range in the main channel and is not used for storing stage data. The storage layout in the main channel is replicated to facilitate subsequent tracking of stage data transfer. Since multiple internal points are scattered throughout the main channel, invalid data is filled between the main channel and the internal points to confuse unauthorized network access. The difficulty for unauthorized network access is: it needs to find the internal points in a large amount of data in the main channel. When accessing the main channel, the invalid data provides a buffer time for data transfer. Even if the internal point is found later, the stage data has already been transferred away. This can effectively protect the stage data in the internal points, prevent the data from being obtained or tampered with by unauthorized networks, and ensure the accuracy of subsequent tracing of wind turbine main shaft faults.

[0069] In one embodiment, step S3, which involves transferring data from the root point when an anomaly occurs in the stage channel, associating the transferred data with the tactile paving, and restoring the stage data in the stage channel through the root point after the stage channel returns to normal, includes:

[0070] S31. The main channel of the corresponding stage channel is set with an access port, and the corresponding access port is marked with an authorized network. When an unauthorized network accesses the main channel, it indicates that there is an anomaly in the stage channel.

[0071] S32. When an anomaly occurs in the stage channel, open the switching port to transfer the mobile point in the inner point and the stage type data of the preset time period to the outer point through the switching port, and close the switching port to disconnect the outer point from the main channel, always maintaining the connection between the mobile point and the corresponding associated point.

[0072] S33. When an unauthorized network exits access, it indicates that the phase channel is normal. The external point is reconnected to the corresponding switching port based on the mobile point. The mobile point in the external point and the phase type data of the preset time period are transferred to the internal point through the switching port to complete the recovery of the phase data in the main channel.

[0073] As described in steps S31-S33 above, an access port is set for the main channel. This access port is a port that can be used by the management port and the system to access and connect when tracing data. The network range where the authorized management port and other access ports are located is designated as the authorized network, and the access port is marked as an authorized network. When an unauthorized network accesses the main channel, it indicates that there is an anomaly in the stage channel. When there is an anomaly in the stage channel, the switching port is opened, and the movement point in the inner point and the stage type data of the preset time period are transferred to the outer point through the switching port. After the transfer, the switching port is closed, and the outer point is disconnected from the main channel. The outer point and the switching port are not just closed, but directly disconnected. There is another storage space outside the main channel. That is to say, the main channel is in one storage space, and the outer point is in a storage space outside the main channel after being disconnected from the switching port. At the same time, the connection between the movement point and the corresponding associated point is always maintained. When the unauthorized network exits access, it indicates that the stage is abnormal. When the main channel is functioning normally, based on the connection between the associated points and the moving points in the tactile paving, the outer point where the moving point is located can be accurately reconnected to the previously disconnected exchange port. Then, the moving point in the outer point and the stage type data of the preset time period can be transferred to the inner point through the exchange port, completing the recovery of the stage data in the main channel. When the main channel is accessed by an unauthorized network, the reception and storage of data will be suspended. Since the operating data of the wind turbine main shaft is continuously generated and collected, a temporary receiving and storage space can be set in the traceability platform or other cloud servers. After recovery, the temporary receiving and storage space is connected to the main channel to load and store the data. The specific data loading and storage method is the same as the storage method of the real-time collected stage data in the main channel. It can securely store the collected operating data of the wind turbine main shaft, ensure the accuracy of the data during the storage process, and ensure the long-term data security. It can be directly used for fault analysis and fault tracing of the wind turbine main shaft.

[0074] In one embodiment, step S4, which involves constructing stage data into a digital twin model based on stage channels for simulation, extracting key data at the time of the fault when a wind turbine main shaft fails, tracing relevant data corresponding to the key data based on stage channels, and analyzing the relevant data to obtain the root cause data of the fault, includes:

[0075] S41. The receiving port in the digital twin model receives stage data and stores it in the corresponding stage channel;

[0076] S42. Based on the operating data of the wind turbine main shaft, a digital twin model is used for simulation. When the wind turbine main shaft fails, key data at the time of the failure is extracted, and relevant data corresponding to the key data are traced based on the stage channel.

[0077] S43. Perform fault correlation analysis on relevant data to obtain fault root cause data.

[0078] As described in steps S41-S43 above, extract key information at the time of the fault, specifically including the fault type, such as wear, fracture, or abnormal noise; the time of occurrence and operating conditions, such as load, speed, and ambient temperature. Associate the entire lifecycle data of the faulty spindle, filtering out relevant subsets of data from each stage of design, manufacturing, and maintenance, and excluding irrelevant and interfering data. The specific operations are as follows: First, clarify the core dimensions and objectives for filtering: first, identify the core attributes of the fault, including the fault type, such as fatigue fracture or excessive wear; the fault location, such as the shaft shoulder or bearing mating surface; and the fault occurrence sequence, such as after T hours of operation or under specific operating conditions. Define the filtering objectives: only retain data that reflects the rationality of the design, manufacturing accuracy, and operational standardization, and is related to the fault attributes, excluding redundant information unrelated to the fault, such as machining records of non-faulty parts and irrelevant environmental data under normal operating conditions.

[0079] The next step is data filtering during the design phase. Key filtering fields include: structural design parameters related to the fault location, such as dimensions, wall thickness, and chamfer radius; material selection data, such as tensile strength and fatigue limit; and simulation data corresponding to the fault conditions, such as stress distribution and life prediction results. Auxiliary design data not related to the fault location, general standard documents, and simulation scenario data unrelated to the fault conditions, such as extreme low-temperature simulations (where the actual fault occurs at room temperature), will be removed. Key filtering fields include: material batch inspection data of the faulty spindle, such as composition content and mechanical properties; machining process parameters of the faulty location, such as cutting speed, heat treatment temperature, and grinding accuracy; and quality inspection data for that location, such as dimensional tolerances, surface roughness, and flaw detection results. Manufacturing data for other components, machining records for non-faulty locations, general production process documents, and equipment operation logs unrelated to the fault, such as workshop lighting data during machining, will be deleted.

[0080] Next, data filtering is performed during the operation and maintenance phase. Core filtering fields include: operational data for a preset number of cycles before the failure, such as vibration amplitude, temperature changes, and load fluctuations; maintenance records for the faulty component, such as lubricant change time and repair results; and real-time operating conditions at the time of the failure, such as speed, load, and ambient humidity. Irrelevant operational data after the failure, maintenance records for other components, and stable operational data under normal conditions, such as data without fluctuations or abnormal warnings, as well as environmental data unrelated to the failure, such as ambient noise far from the spindle, are excluded.

[0081] Finally, a multi-level screening method is used to refine the data: First level: Tag screening, based on predefined tags for traceable data, such as fault location = shoulder, data type = processing parameters, time range = 6 months before the fault. This data can be predefinedly marked in the stage channel to achieve traceability of subsequent data with predefined tags, quickly filtering out completely irrelevant data. Second level: Threshold screening, setting reasonable data ranges, such as standard values ​​for material tensile strength, processing dimensional tolerance ranges, and normal operating vibration thresholds; filtering out abnormal data exceeding the range, retaining deviation data that may be related to the fault. Third level: Correlation screening, using algorithms such as correlation analysis to calculate the correlation between data and fault attributes, eliminating data with a correlation below a set threshold; if the correlation is less than the preset value, it represents non-critical data. Fourth level: Manual review, where the technical team involved in the entire lifecycle manually verifies the screened data to eliminate interference data misjudged by the algorithm, such as occasional abnormal sensor data. The valid data filtered at each stage are managed according to their fault correlation. This correlation can be used to identify data within the stage channel. Data identifiers corresponding to fault correlation are recorded and linked separately to form a structured data subset. This subset includes key information such as data source (e.g., from a sensor collecting operational data on the wind turbine main shaft), timestamp, numerical value, and anomaly description. The data subset is then standardized, with unified units and corrected formats to ensure data consistency during subsequent comparative analysis.

[0082] The mapping relationship between load, rotational speed, and spindle life loss is essentially a quantitative model of life loss under the combined effect of load and rotational speed. The core conclusion is that life loss is non-linearly positively correlated with load and positively correlated with rotational speed, and there is a coupling effect between the two. Specifically, based on the fatigue damage accumulation theory, spindle life loss is the cumulative result of long-term load and rotational speed. Load directly determines the stress level of the spindle; once the stress exceeds the material's fatigue limit, fatigue damage will occur with each cycle. Rotational speed determines the stress cycle frequency; the higher the rotational speed, the faster the damage accumulates per unit time. The coupling effect between the two is as follows: Under high load, increased rotational speed accelerates the stress cycle count and may also amplify vibration, further exacerbating life loss; under low load, the effect of rotational speed on life loss is relatively mild.

[0083] Establishment and verification of mapping relationships: Collect SN curve data of the spindle material, and experimentally test the fatigue life under different load and speed combinations to obtain parameters such as k, m, and p. Combined with a digital twin model, input design parameters and material properties to simulate stress distribution and cyclic characteristics under different load-speed conditions, and optimize the accuracy of the mapping model. Use historical operating data of the physical spindle, such as load and speed time series data, and actual life loss, such as operating years and wear amount, for reverse verification, adjust model parameters, and ensure that the mapping relationship fits reality; Third level: Correlation screening. The correlation analysis algorithm is used to calculate the degree of association between data and fault attributes, and data with a correlation degree below a set threshold, such as non-critical data with a correlation degree <0.3, are removed. The core definition of correlation degree is essentially a quantitative value of the "degree of influence" of data features on core attributes such as fault occurrence, fault type, and fault location. For example, the correlation degree between spindle vibration amplitude data and fatigue fracture fault is 0.8, which is a strong positive correlation, while the correlation degree between workshop lighting data and the same fault is only 0.04, which can be considered as no correlation. The purpose is to use numerical values ​​to clearly define the strength of the correlation between data and faults, avoiding the subjective judgment of human intervention. The process involves precise removal of invalid data. The steps for obtaining correlation are as follows: First, clarify the analysis object and data preprocessing; second, determine the target variable: i.e., fault attributes, such as whether a fault occurred (binary classification: 0 = normal, 1 = fault), fault severity (multi-classification: such as mild / moderate / severe), and fault location (e.g., labeling: shaft shoulder / bearing surface, etc.); third, determine the characteristic variables: i.e., the full lifecycle data to be filtered, such as structural dimensions in the design phase, heat treatment temperature in the manufacturing phase, vibration values ​​in the operation and maintenance phase, etc., which need to be converted into numerical data in advance, such as encoding labeled data as numbers. Data cleaning: Remove missing and outlier values ​​(such as sensor false alarms) from feature variables to ensure calculation accuracy. Select an appropriate correlation calculation algorithm: choose the corresponding algorithm based on the type of fault attribute (target variable). For example, in one scenario: the fault attribute is binary / multi-class, such as whether it is a fault or the type of fault; common algorithms include mutual information, chi-square test, and decision tree feature importance. Measure the degree of information dependence between the feature variable and the fault category; for example, the higher the mutual information value, the more the feature can reflect the differences in fault categories. In another scenario: the fault attribute is continuous, such as the vibration peak value or life loss during a fault. The Pearson coefficient can be used to measure the degree of linear correlation, and the Spearman coefficient can be used to measure the degree of monotonic correlation (suitable for nonlinear relationships). In yet another scenario, it is necessary to clarify the causal relationship, such as whether a certain processing parameter causes the fault; common algorithms include Bayesian networks and propensity score matching; eliminate confounding factors and quantify the direct impact of feature variables on the fault. Design-stage source tracing and comparison: retrieve the design parameters of the fault's main axis, such as structural dimensions, material selection, strength verification standards, and simulation data, such as fatigue life and stress distribution simulation results.Compare the design baseline data of a normal spindle of the same model to analyze whether there are unreasonable design parameters, such as insufficient safety factors or mismatch between simulation boundary conditions and actual working conditions. Verify design review records to determine if there are design flaws or unconsidered extreme working scenarios. Extract manufacturing data from the faulty spindle, including material batch inspection reports, machining parameters such as cutting speed and heat treatment temperature, and quality inspection data such as dimensional tolerances and surface roughness. Compare the manufacturing data range of qualified batch spindles to identify manufacturing deviations such as excessive material composition, machining parameters deviating from standard values, or missed inspections during quality control. Track assembly process records to investigate assembly errors and improper component fit. Perform source tracing and comparison in the operation and maintenance process: analyze the operating data trend of the spindle before the failure, including vibration amplitude, temperature changes, and load fluctuations, to determine if there are long-term overloads or abnormal operating conditions. Verify operation and maintenance records to confirm whether there are untimely maintenance issues, such as failure to change lubricating oil on time, improper maintenance operations such as omissions in the maintenance process, or failure to respond to fault warnings. By comparing the operation and maintenance data of a normal spindle under the same working conditions, the differences in the operation and maintenance process can be identified. Based on the hierarchical comparison of abnormal data, a causal chain between abnormal factors and failure phenomena can be constructed to clarify the correlation strength between each abnormal factor and the failure. Simulation verification using a digital twin model can be conducted to simulate whether the spindle will reproduce the failure under the individual or combined effects of design defects, manufacturing deviations, or improper operation and maintenance. A joint review by the technical team (design, manufacturing, and operation and maintenance) is conducted to eliminate secondary factors and pinpoint the root cause of the failure.

[0084] Example 2, please refer to Figure 2 As shown in this embodiment, the wind turbine main shaft full lifecycle traceability system based on digital twin includes:

[0085] The module is used to collect physical data of the wind turbine main shaft and build a digital twin model of the wind turbine main shaft based on the physical data and performance model;

[0086] The configuration module, connected to the construction module, is used to collect and store the full life cycle data of the wind turbine main shaft into the corresponding stage channel to obtain stage data. Multiple root points are set in the stage channel, and the stage data is bound based on the root points.

[0087] The transfer module, connected to the setting module, is used to transfer the data in the root point when there is an anomaly in the stage channel, and associate the transferred data with the tactile paving. After the stage channel is normal, the stage data in the stage channel is restored through the root point.

[0088] The analysis module, connected to the setting module, is used to build stage data into a digital twin model for simulation based on stage channels. When a fault occurs in the main shaft of the wind turbine, the key data at the time of the fault is extracted, and the relevant data corresponding to the key data is traced based on the stage channels. The relevant data is analyzed to obtain the root cause data of the fault.

[0089] In one embodiment, the building module includes:

[0090] The data acquisition unit is used to construct the basic model based on physical data, which includes geometric data, material mechanical parameters, and thermodynamic properties.

[0091] The fusion unit is used to merge the performance model with the basic model to obtain a digital twin model of the wind turbine main shaft and store it in the traceability platform.

[0092] It should be noted that the stage channel can effectively protect the stage data in the root point, preventing the data from being obtained or tampered with by unauthorized networks. The data of the entire life cycle of the wind turbine main shaft needs to be stored for a long time, so that the data remains secure under long-term storage and can be directly used for fault analysis and fault tracing of the wind turbine main shaft, avoiding incorrect fault tracing of the wind turbine main shaft due to the insecurity of stored data.

[0093] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for tracing the entire lifecycle of a wind turbine main shaft based on digital twins, characterized in that, Includes the following steps: Collect physical data of the wind turbine main shaft, and build a digital twin model of the wind turbine main shaft based on the physical data and performance model; The entire lifecycle data of the wind turbine main shaft is collected and stored in the corresponding stage channel to obtain stage data. Multiple root points are set in the stage channel, and the stage data is bound based on the root points. In the phased passageway, a main passageway and a tactile paving are set up, and the tactile paving and the main passageway are connected. Multiple root points are set in the main channel. The root points include internal points, external points and switching ports. The switching ports are set in the main channel. The internal points and external points are connected through the switching ports. Each internal point includes a moving point. The moving point corresponds to the switching port connected to its internal point. When there is no abnormality in the stage channel, the switching port is in the closed state. When an anomaly occurs in the stage channel, the data in the root point is transferred, and the transferred data is correlated with the tactile paving. After the stage channel returns to normal, the stage data in the stage channel is restored through the root point, including the following steps: The main channel of the corresponding stage channel is configured with an access port, and the corresponding access port is marked with an authorized network. When an unauthorized network accesses the main channel, it indicates that there is an anomaly in the stage channel. When an anomaly occurs in the stage channel, the switching port is opened to transfer the movement point in the inner point and the stage type data of the preset time period to the outer point through the switching port, and the switching port is closed to disconnect the outer point from the main channel, while always maintaining the connection between the movement point and the corresponding associated point. When an unauthorized network exits access, it indicates that the phase channel is normal. The external point is reconnected to the corresponding switching port based on the mobile point. The mobile point and the phase type data of the preset time period in the external point are transferred to the internal point through the switching port to complete the recovery of the phase data in the main channel. Based on the stage channel, stage data is built into a digital twin model for simulation. When the main shaft of the wind turbine fails, key data at the time of the failure is extracted. Based on the stage channel, the relevant data corresponding to the key data is traced, and the relevant data is analyzed to obtain the root cause data of the failure.

2. The method for full lifecycle traceability of wind turbine main shaft based on digital twin as described in claim 1, characterized in that, The steps of collecting physical data of the wind turbine main shaft and building a digital twin model of the wind turbine main shaft based on the physical data and performance model include: The basic model is constructed based on physical data, which includes geometric data, material mechanical parameters, and thermodynamic properties. The performance model is fused with the basic model to obtain a digital twin model of the wind turbine main shaft, which is then stored in the traceability platform.

3. The method for full lifecycle traceability of wind turbine main shaft based on digital twin as described in claim 1, characterized in that, The steps of collecting and storing the full lifecycle data of the wind turbine main shaft into the corresponding stage channel to obtain stage data, setting multiple root points in the stage channel, and binding the stage data based on the root points include: Multiple data collection points are set up for the main shaft of the wind turbine. Based on the data collection points, corresponding receiving ports are set up in the digital twin model, and a communication connection is established between the data collection points and the corresponding receiving ports. In the traceability platform, corresponding stage channels are set up for the entire life cycle stage of the wind turbine main shaft. The operating data of the wind turbine main shaft is collected based on the collection points and marked with timestamps. The operating data marked with timestamps is transmitted to the corresponding receiving port and stored in the stage channel of the traceability platform as stage data until the stage data corresponding to the entire life cycle stage is obtained, which is used as the entire life cycle data of the wind turbine main shaft. The stage data in the stage channel is captured and stored by using the root point.

4. The method for full lifecycle traceability of wind turbine main shaft based on digital twin as described in claim 3, characterized in that, The step of retrieving and storing stage data from the stage channel through the root point includes: A correlation diagram is set up in the tactile paving. The correlation diagram consists of multiple correlation points. Each correlation point is connected to a movement point in a one-to-one correspondence. When the inner point where the movement point is located does not store stage data, the correlation point and the movement point are disconnected. When storing stage data in a stage channel, the stage data is classified based on the data type to obtain multiple stage type data. The storage range of the main channel in the corresponding stage channel is divided according to the data type, and multiple storage type ranges of the corresponding data type are obtained respectively. Based on the timestamp, the stage type data is stored in the corresponding storage type range of the preset time period, thus completing the capture and storage of the stage data. After storing the stage type data of the preset time period in the inlier, the connection relationship between the moving point in the inlier and the corresponding associated point in the association graph is enabled, and the associated points in the association graph are divided and associated according to the inlier corresponding to the storage type range. Fill the gap between the main channel and the interior point with invalid data.

5. The method for full lifecycle traceability of wind turbine main shaft based on digital twin as described in claim 1, characterized in that, The steps of constructing stage data into a digital twin model based on stage channels for simulation, extracting key data at the time of failure when the wind turbine main shaft fails, tracing relevant data corresponding to the key data based on stage channels, and analyzing the relevant data to obtain the root cause data of the failure include: In the digital twin model, the receiving port receives stage data and stores it in the corresponding stage channel; The simulation is driven by the digital twin model based on the operating data of the wind turbine main shaft. When the wind turbine main shaft fails, the key data at the time of the failure is extracted, and the relevant data corresponding to the key data is traced based on the stage channel. Fault correlation analysis was performed on the relevant data to obtain the root cause data of the faults.

6. A digital twin-based wind turbine main shaft lifecycle traceability system, used to implement the digital twin-based wind turbine main shaft lifecycle traceability method according to any one of claims 1-5, characterized in that, include: The module is used to collect physical data of the wind turbine main shaft and build a digital twin model of the wind turbine main shaft based on the physical data and performance model; The configuration module, connected to the construction module, is used to collect and store the full life cycle data of the wind turbine main shaft into the corresponding stage channel to obtain stage data. Multiple root points are set in the stage channel, and the stage data is bound based on the root points. The transfer module, connected to the setting module, is used to transfer the data in the root point when there is an anomaly in the stage channel, and associate the transferred data with the tactile paving. After the stage channel is normal, the stage data in the stage channel is restored through the root point. The analysis module, connected to the setting module, is used to build stage data into a digital twin model for simulation based on stage channels. When a fault occurs in the main shaft of the wind turbine, the key data at the time of the fault is extracted, and the relevant data corresponding to the key data is traced based on the stage channels. The relevant data is analyzed to obtain the root cause data of the fault.

7. The wind turbine main shaft full life cycle traceability system based on digital twin as described in claim 6, characterized in that, The building module includes: The data acquisition unit is used to construct the basic model based on physical data, which includes geometric data, material mechanical parameters, and thermodynamic properties. The fusion unit is used to merge the performance model with the basic model to obtain a digital twin model of the wind turbine main shaft and store it in the traceability platform.

Images