Intelligent early warning method and system for wind turbines
By performing consistent clustering and asynchronous monitoring and configuration of wind turbine clusters, the information redundancy caused by monitoring multiple key components in wind turbine clusters is solved, and efficient early warning analysis and decision support is achieved.
Patent Information
- Application Number
- CN202411290988.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-14
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-09-14
AI Technical Summary
The simultaneous monitoring of multiple key components in the wind turbine cluster leads to redundancy of information, resulting in lag in the early warning system, and the inability to respond to potential faults or abnormal situations in a timely and effective manner.
By obtaining the layout information of the wind turbine cluster, extracting monitoring data according to the preset component monitoring indicators, performing consistency cluster authentication and monitoring cluster division, configuring data processing components, traversing the monitoring cluster for asynchronous monitoring configuration, and generating monitoring and early warning results.
It significantly improves the monitoring and early warning efficiency of wind turbine clusters, reduces redundant data processing, improves the targeted and real-time nature of early warnings, and provides timely and accurate decision-making basis for operation and maintenance personnel.
Smart Images

Figure CN118793575B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of generator management technology, specifically to the field of generator early warning technology, and in particular to an intelligent early warning method and system for wind turbines. Background Art
[0002] With the growing global demand for clean energy, wind power generation, as an important part of renewable energy, has seen rapid expansion in technology development and application. However, with the large-scale, intelligent and widely distributed wind turbines, the operation and maintenance of wind turbines has become increasingly complex, especially in harsh environments and long-term operation conditions, where the monitoring and early warning of key components are particularly important. Wind turbines are composed of many key components, each of which may generate a large amount of monitoring data during operation. If all nodes are monitored simultaneously, information overload will inevitably occur. Processing this data not only consumes a lot of computing resources, but may also cause lags in the early warning system due to the large amount of data, making it impossible to respond to potential faults or abnormal situations in a timely and effective manner. Summary of the invention
[0003] The present application provides an intelligent early warning method and system for wind turbines, aiming to solve the technical problem that simultaneous monitoring of multiple key components in a wind turbine cluster causes information redundancy, resulting in a strong lag in early warning.
[0004] In view of the above problems, the present application provides an intelligent early warning method and system for wind turbines.
[0005] The first aspect disclosed in the present application provides an intelligent early warning method for wind turbines, the method comprising: obtaining cluster layout information of a target wind turbine cluster, wherein the cluster layout information includes a plurality of wind turbines and a plurality of wind turbine layout positions; extracting monitoring values for a plurality of key component sets of the plurality of wind turbines according to a preset component monitoring indicator set at a current moment, generating a plurality of key component monitoring data sets; performing consistency clustering authentication on the plurality of key component monitoring data sets, and dividing the plurality of wind turbines into monitoring clusters according to the authentication results, obtaining K monitoring clusters, wherein the K monitoring clusters include K clustered key component monitoring data sets and K divided wind turbine sets; according to The K clustered key component monitoring data sets, the K divided wind turbine sets and the multiple wind turbine layout locations are used to configure K data processing components of the data processing unit, wherein the K data processing components correspond one-to-one to the K monitoring clusters; the K clustered key component monitoring data sets are traversed to perform asynchronous monitoring configuration of key components, and K key component asynchronous monitoring schemes of the K monitoring clusters are obtained; based on the K key component asynchronous monitoring schemes, key component monitoring data are collected for the K divided wind turbine sets to obtain K key component asynchronous monitoring data sets, and the K key component asynchronous monitoring data sets are input into the K data processing components for early warning analysis to obtain multiple monitoring and early warning results.
[0006] Another aspect disclosed in the present application provides an intelligent early warning system for wind turbines, the system comprising: a layout information acquisition module, the layout information acquisition module is used to obtain cluster layout information of a target wind turbine cluster, wherein the cluster layout information includes multiple wind turbines and multiple wind turbine layout positions; a monitoring value extraction module, the monitoring value extraction module is used to extract monitoring values for multiple key component sets of the multiple wind turbines according to a preset component monitoring indicator set at a current moment, and generate multiple key component monitoring data sets; a monitoring cluster division module, the monitoring cluster division module is used to perform consistency clustering authentication on the multiple key component monitoring data sets, and divide the multiple wind turbines into monitoring clusters according to the authentication results to obtain K monitoring clusters, wherein the K monitoring clusters include K clustered key component monitoring data sets and K divided wind turbine sets; a data processing unit configuration Module, the data processing unit configuration module is used to configure K data processing components of the data processing unit according to the K clustered key component monitoring data sets, the K divided wind turbine sets and the layout positions of the multiple wind turbines, wherein the K data processing components correspond to the K monitoring clusters one by one; a key component asynchronous monitoring configuration module, the key component asynchronous monitoring configuration module is used to traverse the K clustered key component monitoring data sets to perform key component asynchronous monitoring configuration, and obtain K key component asynchronous monitoring schemes for the K monitoring clusters; an early warning analysis module, the early warning analysis module is used to collect key component monitoring data for the K divided wind turbine sets based on the K key component asynchronous monitoring schemes, obtain K key component asynchronous monitoring data sets, and input the K key component asynchronous monitoring data sets into the K data processing components for early warning analysis to obtain multiple monitoring and early warning results.
[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:
[0008] The above-mentioned intelligent early warning method for wind turbines collects the layout information of the entire cluster, including the specific location of each generator. At the current moment, according to the preset monitoring indicators, the monitoring data of the key components of each wind turbine are extracted. These data are used for consistency clustering analysis, the purpose of which is to classify similar wind turbines into the same monitoring cluster to reduce redundant data processing and improve the pertinence of early warning. Through clustering, K monitoring clusters are obtained, each of which contains a set of similar monitoring data of wind turbines and their key components. Subsequently, data processing components are configured for the K monitoring clusters to ensure that the data of each cluster can be processed efficiently. After that, an asynchronous monitoring scheme is customized for each monitoring cluster, so that the key components of different clusters can be independently and efficiently monitored according to their characteristics and requirements. Based on these asynchronous monitoring schemes, the monitoring data of the key components of each monitoring cluster are collected, and these data are sent to the corresponding data processing components for analysis. Then, each data processing component will generate monitoring and early warning results, which reflect the health status and potential risks of wind turbines in each monitoring cluster, providing timely and accurate decision-making basis for operation and maintenance personnel. The entire process significantly improved the monitoring and early warning efficiency of wind turbine clusters by optimizing monitoring resource allocation and adopting asynchronous monitoring strategies.
[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0011] Figure 1 The figure is a flow chart of an intelligent early warning method for a wind turbine in one embodiment.
[0012] Figure 2 The figure is an architecture diagram of an intelligent early warning system for wind turbines in one embodiment.
[0013] Explanation of the accompanying drawings: deployment information acquisition module 1, monitoring value extraction module 2, monitoring cluster division module 3, data processing unit configuration module 4, key component asynchronous monitoring configuration module 5, early warning analysis module 6. DETAILED DESCRIPTION
[0014] The embodiments of the present application provide an intelligent early warning method and system for wind turbines to solve the technical problem that simultaneous monitoring of multiple key components in a wind turbine cluster causes information redundancy, resulting in a strong lag in early warning.
[0015] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.
[0016] It should be noted that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.
[0017] Embodiment 1, as Figure 1 As shown, the present application provides an intelligent early warning method for a wind turbine, the method comprising:
[0018] The cluster layout information of the target wind turbine cluster is obtained, wherein the cluster layout information includes a plurality of wind turbines and a plurality of wind turbine layout positions.
[0019] The intelligent early warning method for wind turbines and its corresponding intelligent early warning system play a vital role in the field of modern wind power generation. The intelligent early warning system realizes efficient monitoring of the operating status of wind turbines and fault early warning by communicating with the data processing unit.
[0020] In the embodiment of the present application, in the management of the wind turbine cluster, the system terminal first obtains the layout of the target wind turbine cluster. This includes understanding how many wind turbines are included in the cluster and the specific location of each turbine in the geographic space. This information constitutes the core content of the cluster layout information, which is crucial for subsequent monitoring and early warning.
[0021] At the current moment, according to the preset component monitoring index set, monitoring values are extracted for the multiple key component sets of the multiple wind turbines respectively to generate multiple key component monitoring data sets.
[0022] In one embodiment, at the current time point, the system terminal monitors multiple key components of each generator in the wind turbine cluster based on a series of pre-set monitoring indicators of key components. These monitoring indicators include key parameters such as temperature, vibration, current, and voltage. By monitoring these components in real time, the system terminal extracts their current monitoring values and organizes these values into a set of monitoring data for multiple key components. In this way, the system terminal obtains real-time status data of the key components of each generator, providing basic data support for subsequent analysis and early warning.
[0023] Furthermore, the present application provides a set of preset component monitoring indicators, including:
[0024] Obtain blade monitoring indicators, wherein the blade monitoring indicators include blade vibration, stress and blade temperature; obtain gearbox monitoring indicators, wherein the gearbox monitoring indicators include oil temperature, noise and box vibration; obtain generator monitoring indicators, wherein the generator monitoring indicators include current, voltage, operating temperature and frequency; use the blade monitoring indicators, gearbox monitoring indicators and generator monitoring indicators as the preset component monitoring indicator set.
[0025] Preferably, when monitoring a wind turbine cluster, the system terminal obtains monitoring indicators of multiple key components, which are set based on actual needs and professional knowledge. Among them, the multiple key components include blades, gearboxes, generators, etc. For blades, the system terminal obtains blade monitoring indicators, including blade vibration, stress, and blade temperature. Among them, blade vibration describes the vibration state generated by the wind turbine blades during rotation, and is an important indicator for evaluating the structural integrity and performance of the blades. Stress is used to describe the internal resistance of an object when subjected to external forces. Blade temperature refers to the temperature on the surface of the wind turbine blade, which may be affected by multiple factors such as ambient temperature, wind speed, and sunshine. It is an important basis for evaluating the working state and potential problems (such as overheating) of the blades. For gearboxes, the system terminal obtains gearbox monitoring indicators, including oil temperature, noise, and housing vibration. Among them, oil temperature refers to the temperature of the lubricating oil in the gearbox. Excessive oil temperature may indicate problems such as overheating, poor lubrication, or cooling system failure. Noise describes the sound generated by the wind turbine during operation. Abnormal noise may indicate wear, imbalance, or other potential problems of mechanical parts. Housing vibration refers to the vibration generated by the gearbox housing during operation. Excessive vibration levels may indicate a fault or imbalance problem inside the component. For the generator, the system terminal obtains the generator monitoring indicators, including current, voltage, operating temperature, frequency, etc. Among them, the current describes the rate and direction of charge flow in the generator, and is an important indicator for evaluating the operating status and performance of the equipment. The voltage describes the magnitude of the potential difference in the electric field and is a basic parameter in the transmission and use of electric energy. Stable voltage is essential to ensure the normal operation of the generator. The operating temperature describes the temperature of the generator during normal operation. Excessive operating temperature may indicate that the generator is overloaded. The frequency describes the number of periodic changes of alternating current per unit time and is one of the important indicators of power quality. The frequency of the alternating current output by the generator needs to be stable near the rated value to ensure the stable operation of the generator. The system terminal combines the above monitoring indicators for blades, gearboxes and generators to form a preset component monitoring indicator set for wind turbine monitoring. This set comprehensively covers the main components and key parameters of wind turbines, providing a comprehensive basis for the subsequent monitoring value extraction.
[0026] The plurality of key component monitoring data sets are subjected to consistency clustering authentication, and the plurality of wind turbines are divided into monitoring clusters according to the authentication results to obtain K monitoring clusters, wherein the K monitoring clusters include K clustered key component monitoring data sets and K divided wind turbine sets.
[0027] In one embodiment, the system terminal analyzes the monitoring data set of multiple key components, and identifies the similarities and differences in the data through consistency clustering authentication combined with the consistency clustering network layer. Based on this authentication result, multiple wind turbines are effectively divided into K monitoring clusters. The wind turbines in each monitoring cluster show a high degree of similarity in the monitoring data of key components, that is, their operating status and performance characteristics are relatively consistent. This division method helps to manage the wind turbine cluster more finely, because each monitoring cluster can be regarded as a subset with similar characteristics. For each monitoring cluster, a more accurate data processing component can be configured to improve the operating efficiency and reliability of the entire cluster. At the same time, since the status of the wind turbines in the monitoring cluster is relatively consistent, the redundancy of monitoring information and the problem of early warning lag caused by individual differences are also reduced.
[0028] Furthermore, the present application provides a method of obtaining the K divided wind turbine generator sets, comprising:
[0029] K key component monitoring data sets are randomly extracted from the multiple key component monitoring data sets to obtain a first extraction result, wherein K is a positive integer greater than or equal to 3; the K key component monitoring data sets are enumerated in pairs to obtain multiple enumeration combinations; the multiple enumeration combinations are traversed to perform collaborative consistency analysis to obtain multiple collaborative consistency factors.
[0030] Preferably, the system terminal selects K data sets from multiple key component monitoring data sets by random sampling as the first extraction result, wherein K is a positive integer greater than or equal to 3. This step is intended to ensure the diversity and representativeness of the sample through randomness. Subsequently, the K key component monitoring data sets are enumerated in pairs to generate multiple enumeration combinations. Each enumeration combination contains two data sets, and these combinations provide a basis for subsequent analysis. For each pair of two data sets in the combination, the system terminal calculates the Pearson correlation coefficient between the corresponding features in the two data sets. This is achieved by performing statistical analysis on the samples in the data set to obtain the correlation coefficient of each feature, and then aggregating the correlation coefficients of all features, that is, calculating the average value of the correlation coefficients of all features to obtain the collaborative consistency factor. This analysis process is intended to evaluate the similarity between the two data sets and quantify the collaborative consistency factor. The smaller the collaborative consistency factor, the better, because the partition nodes in the decision tree will be constructed based on the extracted multiple data sets in the future. Low similarity means that the differences between the data sets are obvious, which helps to form a more effective partition in the decision tree, thereby improving the accuracy of classification. Therefore, the calculation of the collaborative consistency factor is an important step before building a decision tree to divide nodes, which provides a key basis for subsequent data analysis and network layer construction. By ensuring low similarity between data sets, the decision tree can more accurately reflect the internal relationships and laws between data.
[0031] A consistency clustering network layer is constructed based on the multiple collaborative consistency factors and the K key component monitoring data sets of the first extraction results; the multiple key component monitoring data sets are input into the consistency clustering network layer to generate K clustered key component monitoring data sets; based on the mapping relationship between the key component monitoring data sets and the wind turbines, the K clustered key component monitoring data sets are combined to obtain the K divided wind turbine sets.
[0032] Preferably, the system terminal constructs a consistency clustering network layer based on the calculated multiple collaborative consistency factors and K data sets randomly selected from multiple key component monitoring data sets, combined with a decision tree. The role of this consistency clustering network layer is to cluster them according to the similarity between the data sets. The lower the collaborative consistency factor, the greater the difference between the two data sets, which helps to separate them in the clustering process. Subsequently, the system terminal inputs all key component monitoring data sets into this consistency clustering network layer. Through the processing of the consistency clustering network layer, these data sets are divided into K clusters, each cluster containing a set of key component monitoring data sets with high similarity. These clustered data sets are defined as K clustered key component monitoring data sets. Afterwards, the mapping relationship between the key component monitoring data set and the wind turbine is used to associate the data in each clustered key component monitoring data set with the corresponding wind turbine. In this way, the wind turbine can also be divided into K sets according to the clustering results to obtain K divided wind turbine sets, and the wind turbines in each divided wind turbine set show high similarity in the monitoring data of the key components. This process realizes the intelligent grouping of wind turbine clusters, providing convenience for subsequent wind turbine monitoring, management and maintenance.
[0033] Furthermore, the present application provides a method for generating a consistent clustering network layer, including:
[0034] The consistency of the first extraction result is authenticated based on the multiple collaborative consistency factors. If the authentication is passed, a clustering network layer is constructed based on the K key component monitoring data sets to generate a consistent clustering network layer. If the authentication fails, a fine-tuning scale is determined based on the consistency authentication result, and the first extraction result is iteratively adjusted according to the fine-tuning scale until the authentication is passed to generate the consistent clustering network layer.
[0035] Optionally, the system terminal performs consistency authentication on the first extraction result based on multiple collaborative consistency factors, that is, compares the calculated multiple collaborative consistency factors with the authentication threshold. If these collaborative consistency factors are all less than or equal to the authentication threshold, it means that the consistency authentication is passed, indicating that these data sets meet the requirements for building a clustering network layer in terms of similarity. Subsequently, the system terminal uses each key component monitoring data set as an initial partition node in the decision tree. This means that the initial form of the decision tree is a tree containing K leaf nodes, each leaf node represents a key component monitoring data set. Then select cosine similarity as a metric to evaluate the similarity between different data sets. Cosine similarity can effectively capture the similarity of data sets in direction, not just size or amplitude. After that, start from the root node (or initial leaf node, if the decision tree has not been further divided), and traverse the decision tree. For each node (that is, each key component monitoring data set), calculate the cosine similarity of other data sets with the node. The data set with a higher similarity to the node, that is, greater than or equal to the preset similarity threshold, is divided under the node to form a child node or update the data set of the node (if the node is allowed to contain multiple data sets). Repeat the above process until all data sets are divided into corresponding nodes or the number of nodes reaches the preset upper limit. After data division, each node (or leaf node) of the decision tree represents a monitoring cluster, which contains data sets with high similarity to the node. Then, the system terminal uses the constructed decision tree as the consistency clustering network layer. If the consistency certification fails, that is, there is a collaborative consistency factor greater than the certification threshold, it means that the initial K key component monitoring data sets extracted do not meet the requirements for building the clustering network layer in terms of similarity. At this time, the system terminal checks which key component monitoring data sets have similarities that do not meet the requirements of the certification threshold, and evaluates the possible impact of these deviations on the construction of the clustering network layer. Then, according to the analysis results, set clear fine-tuning goals, that is, reduce the similarity between which data sets. The system terminal sets the fine-tuning scale based on the similarity difference between the data sets, that is, the number of key component monitoring data sets that need to be re-extracted and updated in a single adjustment. For example, for data sets with greater similarity, set a larger adjustment range or a more frequent adjustment frequency. After determining the fine-tuning scale, the system terminal iteratively adjusts the first extraction result according to the fine-tuning scale, re-extracts and updates the data set until the consistency certification is passed, and uses the same construction method as above to generate a consistent clustering network layer. This process ensures that the final constructed consistent clustering network layer is based on a highly similar and consistent data set.
[0036] According to the K clustered key component monitoring data sets, the K divided wind turbine sets and the multiple wind turbine layout locations, the K data processing components of the data processing unit are configured, wherein the K data processing components correspond one-to-one to the K monitoring clusters.
[0037] In one embodiment, when configuring the data processing unit, the system terminal divides the data processing unit into K independent data processing components based on the results of the K clustered key component monitoring data sets and the transmission distance between each wind turbine and the data processing unit. These K data processing components form a one-to-one correspondence with the K monitoring clusters obtained through consistency authentication and cluster analysis. Each data processing component is used to process the monitoring data of multiple wind turbines in its corresponding monitoring cluster. This configuration ensures the efficiency and pertinence of data processing, because each component is only responsible for processing the data of the wind turbine associated with it, thereby avoiding confusion and redundancy in data processing and improving the speed and accuracy of data processing.
[0038] Furthermore, the present application provides a method for constructing the K data processing components, including:
[0039] According to the K clustered key component monitoring data sets and the plurality of wind turbine layout locations, K wind turbine layout location sets are matched to obtain; based on the K wind turbine layout location sets and the layout location of the data processing unit, K transmission distance sets are determined; the number of wind turbines in the K divided wind turbine sets is counted respectively to obtain the number of K wind turbines; processing resources are configured based on the K transmission distance sets and the number of K wind turbines to obtain K processing allocation resources; and the K data processing components are constructed based on the K processing allocation resources.
[0040] Preferably, the system terminal matches the monitoring data set with the corresponding wind turbine position based on the K clustered key component monitoring data sets and the actual layout positions of the wind turbines, forming K wind turbine layout position sets. These wind turbine layout position sets represent K different groups, and the wind turbines in each group have similar monitoring data characteristics. Subsequently, using these wind turbine layout position sets and the layout positions of the data processing units, K transmission distance sets are calculated. These distance sets reflect the data transmission distance from the data processing unit to the wind turbines in each monitoring area, and are an important basis for evaluating the efficiency and cost of data transmission. Afterwards, the number of wind turbines in each group is counted to obtain the number of K wind turbines. This quantity information helps to understand the scale of each group, thereby providing basic data for subsequent resource allocation. On the basis of obtaining the transmission distance and the number of wind turbines, the system terminal optimizes the configuration of processing resources. By comprehensively considering the efficiency and cost of data transmission, as well as the demand for processing resources of each group, appropriate processing resources are allocated to the K groups, i.e., K processing allocation resources. Then, based on these K processing allocation resources, the data processing unit is divided into K data processing components. These components correspond to the K groups one by one and are responsible for processing the monitoring data of the wind turbines in the corresponding groups. Through such a configuration, refined management and efficient coordination of data processing are achieved, ensuring the stability and reliability of the data processing unit.
[0041] The K clustered key component monitoring data sets are traversed to perform key component asynchronous monitoring configuration, and K key component asynchronous monitoring schemes of the K monitoring clusters are obtained.
[0042] In one embodiment, the system terminal traverses and analyzes K clustered key component monitoring data sets. The purpose of the traversal is to configure asynchronous monitoring for the key components in each monitoring cluster. Specifically, the system terminal formulates a set of special asynchronous monitoring schemes for the key components in each monitoring cluster according to the volatility of its monitoring data. The design of this asynchronous monitoring scheme is intended to ensure that the monitoring of key components can be carried out efficiently and orderly, while avoiding system resource constraints or data delays caused by over-concentration of monitoring tasks. By traversing all K clustered key component monitoring data sets and customizing corresponding asynchronous monitoring schemes for each monitoring cluster, the system terminal will eventually obtain the asynchronous monitoring schemes for the key components of each of the K monitoring clusters. These schemes will serve as a guide for subsequent monitoring work to ensure that each key component can be monitored in a timely and accurate manner.
[0043] Furthermore, the present application provides a method for generating an asynchronous monitoring scheme for the K key components, including:
[0044] Taking key components as indexes, perform fluctuation analysis on the K clustered key component monitoring data sets respectively to determine K key component fluctuation factor sets; perform key component monitoring frequency configuration based on the K key component fluctuation factor sets to generate K component monitoring frequency sets; perform asynchronous monitoring configuration based on the K component monitoring frequency sets to generate the K key component asynchronous monitoring schemes.
[0045] Preferably, the system terminal traverses the K clustered key component monitoring data sets, and uses the key components as indexes to perform monitoring data fluctuation analysis respectively. The core of this analysis is to calculate the fluctuation factor of the same key component, that is, the variance, to evaluate whether the vibration conditions of the blades divided in the same wind turbine set are consistent. Specifically, the system terminal obtains K sets of fluctuation factors of the same key component by calculating the variance of the blade vibration data in each divided wind turbine set. These variance values reflect the discrete degree of the blade vibration conditions in each set, that is, the fluctuation size. If the blade vibration variance in a certain set is small, it means that the vibration conditions of these blades are relatively consistent and may be affected by similar environments or operating conditions. Based on these fluctuation factor sets, we configure the monitoring frequency of key components and obtain K sets of component monitoring frequencies. For blade sets with relatively consistent vibration conditions, the system terminal adopts a wider monitoring frequency because the vibration characteristics of these blades are relatively stable and do not require too frequent monitoring. For blade sets with large vibration fluctuations, a higher monitoring frequency is configured to ensure that potential faults or abnormal conditions can be discovered and handled in a timely manner. Subsequently, the system terminal generates asynchronous monitoring schemes for K key components according to the configuration results of the monitoring frequency, in ascending order of monitoring frequency. These schemes not only take into account the consistency of blade vibration and other conditions within each set, but also take into account the optimization of monitoring efficiency and resource utilization, providing strong support for the stable operation of wind turbines.
[0046] Furthermore, the present application provides a method for generating an asynchronous monitoring scheme for the K key components, including:
[0047] Extract K first component monitoring frequencies from the K component monitoring frequency sets respectively; take the K first component monitoring frequencies as K comparison objects, and use the component monitoring frequencies in the K component monitoring frequency sets whose similarities with the K comparison objects are outside a preset similarity threshold as K first initial synchronization component monitoring frequency sets; use the K first key components corresponding to the K first component monitoring frequencies and the K first initial synchronization key components corresponding to the K first initial synchronization component monitoring frequency sets as K first initial synchronization monitoring component combinations; remove the component monitoring frequencies corresponding to the K first initial synchronization monitoring component combinations from the K component monitoring frequency sets to obtain K sequential monitoring frequency sets; sort the K sequential key component sets corresponding to the K sequential monitoring frequency sets in the order of monitoring frequency from small to large to obtain K first sequential monitoring component sequences; generate the K key component asynchronous monitoring schemes according to the K first initial synchronization monitoring component combinations and the K first sequential monitoring component sequences.
[0048] Optionally, the system terminal extracts K representative first component monitoring frequencies from the K component monitoring frequency sets as comparison objects. These first component monitoring frequencies represent the monitoring frequencies of different key components. Subsequently, the absolute frequency difference between these first component monitoring frequencies and the monitoring frequencies of other components in the set is calculated as the similarity, and then the component monitoring frequencies whose similarity to the comparison object is lower than the preset threshold are extracted to form K first initial synchronization component monitoring frequency sets. The synchronization components in each first initial synchronization component monitoring frequency set do not mean that they all start monitoring at the same time, but although the monitoring frequencies of these components are slightly different, they start at the same time at the beginning, and the probability of overlap of subsequent monitoring data is small, so they can be regarded as a quasi-synchronous monitoring to reduce data conflicts. Afterwards, the key components corresponding to the first component monitoring frequency are combined with the key components corresponding to the first initial synchronization component monitoring frequency set to form K first initial synchronization monitoring component combinations. The components in these combinations will be monitored relatively intensively in the early stage of monitoring, but due to the difference in monitoring frequency, subsequent data processing will not cause too much conflict. Then, the component monitoring frequencies that are already included in the synchronous monitoring combination are eliminated from the K component monitoring frequency sets to obtain K sequential monitoring frequency sets. The component monitoring frequencies in these sets will be monitored in a certain order to avoid data processing conflicts caused by too much monitoring data in the same period. Then, according to the size of the component monitoring frequencies in the sequential monitoring frequency set, the corresponding key component sets are sorted to obtain K first-order monitoring component sequences. This step ensures that the monitoring task can be carried out in an orderly manner from low frequency to high frequency. Finally, the K first initial synchronous monitoring component combinations and the K first-order monitoring component sequences are combined to generate a complete asynchronous monitoring scheme for K key components. This scheme not only takes into account the similarity of monitoring frequencies to reduce data conflicts, but also ensures the orderliness and efficiency of the monitoring task.
[0049] Based on the K key component asynchronous monitoring schemes, key component monitoring data of the K divided wind turbine sets are collected to obtain K key component asynchronous monitoring data sets, and the K key component asynchronous monitoring data sets are input into the K data processing components for early warning analysis to obtain multiple monitoring and early warning results.
[0050] In one embodiment, the system terminal performs efficient and orderly key component monitoring data collection for each subset of the wind turbine set based on the generated K key component asynchronous monitoring scheme. The core of this scheme is to avoid multiple key components from generating a large amount of monitoring data at the same time in the same time period through reasonable monitoring frequency arrangement, thereby reducing the burden of the data processing component and reducing the possibility of data conflict. In the actual monitoring process, the system terminal collects data for the key components in each wind turbine set according to the monitoring frequency and sequence set by the asynchronous monitoring scheme. Since the monitoring is performed asynchronously, each data processing component can focus on processing the currently received monitoring data without too much interference. The collected K key component asynchronous monitoring data sets are then input into the corresponding data processing components. These components have powerful data processing and early warning analysis capabilities, can analyze monitoring data in real time, identify potential faults or abnormal conditions, and generate corresponding monitoring early warning results. Through this process, not only can the real-time monitoring of key components of wind turbines be achieved, but also potential problems can be discovered and warned in time, thereby improving the reliability and operation efficiency of wind turbines.
[0051] In summary, the embodiments of the present application have at least the following technical effects:
[0052] The embodiment of the present application obtains the layout information of the wind turbine cluster, including the location and quantity. Then, according to the preset monitoring indicators, data is collected for the key components of each wind turbine to generate a monitoring data set. After cluster analysis of these data, the wind turbines are divided into several monitoring clusters. According to the data and location of these clusters, the corresponding data processing components are configured. Subsequently, the monitoring data of each cluster is asynchronously monitored and configured, and a monitoring plan is formulated. After that, data collection and early warning analysis are performed based on these plans to generate early warning results to ensure that the key components of the wind turbine are effectively monitored and warned. These technical effects jointly solve the technical problem that the simultaneous monitoring of multiple key components in the wind turbine cluster causes information redundancy, resulting in a strong lag in the early warning, and achieve the effect of reducing the redundancy of monitoring information and improving the real-time and accuracy of early warning analysis by asynchronously monitoring the key components in each monitoring cluster.
[0053] Embodiment 2, based on the same inventive concept as the intelligent early warning method for wind turbines in the above embodiments, Figure 2As shown, the present application provides an intelligent early warning system for wind turbines, the system comprising: a layout information acquisition module 1: the layout information acquisition module 1 is used to obtain cluster layout information of a target wind turbine cluster, wherein the cluster layout information includes multiple wind turbines and multiple wind turbine layout positions; a monitoring value extraction module 2: the monitoring value extraction module 2 is used to extract monitoring values for multiple key component sets of the multiple wind turbines according to a preset component monitoring index set at the current moment, and generate multiple key component monitoring data sets; a monitoring cluster division module 3: the monitoring cluster division module 3 is used to perform consistency clustering authentication on the multiple key component monitoring data sets, and perform monitoring cluster division on the multiple wind turbines according to the authentication results to obtain K monitoring clusters, wherein the K monitoring clusters include K clustered key component monitoring data sets and K divided wind turbine sets; a data processing unit configuration module 4 : The data processing unit configuration module 4 is used to configure the K data processing components of the data processing unit according to the K clustered key component monitoring data sets, the K divided wind turbine sets and the layout positions of the multiple wind turbines, wherein the K data processing components correspond one-to-one to the K monitoring clusters; key component asynchronous monitoring configuration module 5: the key component asynchronous monitoring configuration module 5 is used to traverse the K clustered key component monitoring data sets to perform key component asynchronous monitoring configuration, and obtain K key component asynchronous monitoring schemes for the K monitoring clusters; early warning analysis module 6: the early warning analysis module 6 is used to collect key component monitoring data for the K divided wind turbine sets based on the K key component asynchronous monitoring schemes, obtain K key component asynchronous monitoring data sets, and input the K key component asynchronous monitoring data sets into the K data processing components for early warning analysis to obtain multiple monitoring and early warning results.
[0054] Furthermore, the monitoring value extraction module 2 is also used to execute the following method:
[0055] Obtain blade monitoring indicators, wherein the blade monitoring indicators include blade vibration, stress and blade temperature; obtain gearbox monitoring indicators, wherein the gearbox monitoring indicators include oil temperature, noise and box vibration; obtain generator monitoring indicators, wherein the generator monitoring indicators include current, voltage, operating temperature and frequency; use the blade monitoring indicators, gearbox monitoring indicators and generator monitoring indicators as the preset component monitoring indicator set.
[0056] Furthermore, the monitoring cluster division module 3 is also used to execute the following method:
[0057] K key component monitoring data sets are randomly extracted from the multiple key component monitoring data sets to obtain a first extraction result, wherein K is a positive integer greater than or equal to 3; the K key component monitoring data sets are enumerated in pairs to obtain multiple enumeration combinations; the multiple enumeration combinations are traversed to perform collaborative consistency analysis to obtain multiple collaborative consistency factors; a consistent clustering network layer is constructed based on the multiple collaborative consistency factors and the K key component monitoring data sets of the first extraction result; the multiple key component monitoring data sets are input into the consistent clustering network layer to generate K clustered key component monitoring data sets; based on the mapping relationship between the key component monitoring data sets and the wind turbines, the K divided wind turbine sets are obtained in combination with the K clustered key component monitoring data sets.
[0058] Furthermore, the monitoring cluster division module 3 is also used to execute the following method:
[0059] The consistency of the first extraction result is authenticated based on the multiple collaborative consistency factors. If the authentication is passed, a clustering network layer is constructed based on the K key component monitoring data sets to generate a consistent clustering network layer. If the authentication fails, a fine-tuning scale is determined based on the consistency authentication result, and the first extraction result is iteratively adjusted according to the fine-tuning scale until the authentication is passed to generate the consistent clustering network layer.
[0060] Furthermore, the data processing unit configuration module 4 is also used to execute the following method:
[0061] According to the K clustered key component monitoring data sets and the plurality of wind turbine layout locations, K wind turbine layout location sets are matched to obtain; based on the K wind turbine layout location sets and the layout location of the data processing unit, K transmission distance sets are determined; the number of wind turbines in the K divided wind turbine sets is counted respectively to obtain the number of K wind turbines; processing resources are configured based on the K transmission distance sets and the number of K wind turbines to obtain K processing allocation resources; and the K data processing components are constructed based on the K processing allocation resources.
[0062] Furthermore, the key component asynchronous monitoring configuration module 5 is also used to execute the following method:
[0063] Taking key components as indexes, perform fluctuation analysis on the K clustered key component monitoring data sets respectively to determine K key component fluctuation factor sets; perform key component monitoring frequency configuration based on the K key component fluctuation factor sets to generate K component monitoring frequency sets; perform asynchronous monitoring configuration based on the K component monitoring frequency sets to generate the K key component asynchronous monitoring schemes.
[0064] Furthermore, the key component asynchronous monitoring configuration module 5 is also used to execute the following method:
[0065] Extract K first component monitoring frequencies from the K component monitoring frequency sets respectively; take the K first component monitoring frequencies as K comparison objects, and use the component monitoring frequencies in the K component monitoring frequency sets whose similarities with the K comparison objects are outside a preset similarity threshold as K first initial synchronization component monitoring frequency sets; use the K first key components corresponding to the K first component monitoring frequencies and the K first initial synchronization key components corresponding to the K first initial synchronization component monitoring frequency sets as K first initial synchronization monitoring component combinations; remove the component monitoring frequencies corresponding to the K first initial synchronization monitoring component combinations from the K component monitoring frequency sets to obtain K sequential monitoring frequency sets; sort the K sequential key component sets corresponding to the K sequential monitoring frequency sets in the order of monitoring frequency from small to large to obtain K first sequential monitoring component sequences; generate the K key component asynchronous monitoring schemes according to the K first initial synchronization monitoring component combinations and the K first sequential monitoring component sequences.
[0066] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0067] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.
[0068] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.
Claims
1. An intelligent early warning method for a wind turbine, characterized in that: The method is applied to an intelligent early warning system, the intelligent early warning system is communicatively connected with a data processing unit, and the method comprises: Obtaining cluster layout information of a target wind turbine cluster, wherein the cluster layout information includes a plurality of wind turbines and a plurality of wind turbine layout positions; At the current moment, according to the preset component monitoring indicator set, respectively extract monitoring values of the multiple key component sets of the multiple wind turbines to generate multiple key component monitoring data sets; Performing consistency clustering authentication on the multiple key component monitoring data sets, and dividing the multiple wind turbines into monitoring clusters according to the authentication results to obtain K monitoring clusters, wherein the K monitoring clusters include K clustered key component monitoring data sets and K divided wind turbine sets; According to the K clustered key component monitoring data sets, the K divided wind turbine sets and the plurality of wind turbine layout locations, K data processing components of the data processing unit are configured, wherein the K data processing components correspond one-to-one to the K monitoring clusters; Traversing the K clustered key component monitoring data sets to perform key component asynchronous monitoring configuration, and obtaining K key component asynchronous monitoring solutions of the K monitoring clusters; Based on the K key component asynchronous monitoring schemes, key component monitoring data of the K divided wind turbine sets are collected to obtain K key component asynchronous monitoring data sets, and the K key component asynchronous monitoring data sets are input into the K data processing components for early warning analysis to obtain multiple monitoring and early warning results.
2. The intelligent early warning method for wind turbines according to claim 1, characterized in that: include: Acquiring blade monitoring indicators, wherein the blade monitoring indicators include blade vibration, stress and blade temperature; Acquire gearbox monitoring indicators, wherein the gearbox monitoring indicators include oil temperature, noise and box vibration; Acquire generator monitoring indicators, wherein the generator monitoring indicators include current, voltage, operating temperature and frequency; The blade monitoring index, gearbox monitoring index and generator monitoring index are used as the preset component monitoring index set.
3. The intelligent early warning method for wind turbines according to claim 1, characterized in that: include: Randomly extracting K key component monitoring data sets from the multiple key component monitoring data sets to obtain a first extraction result, where K is a positive integer greater than or equal to 3; Enumerate the K key component monitoring data sets in pairs to obtain multiple enumeration combinations; Traversing the multiple enumeration combinations to perform collaborative consistency analysis to obtain multiple collaborative consistency factors; Constructing a consistency clustering network layer according to the multiple collaborative consistency factors and the K key component monitoring data sets of the first extraction result; Inputting the plurality of key component monitoring data sets into the consistency clustering network layer to generate K clustered key component monitoring data sets; According to the mapping relationship between the key component monitoring data set and the wind turbine, the K clustered key component monitoring data sets are combined to obtain the K divided wind turbine sets.
4. The intelligent early warning method for wind turbines according to claim 3, characterized in that: include: Performing consistency authentication of the first extraction result based on the multiple collaborative consistency factors, and if the authentication is passed, constructing a clustering network layer based on the K key component monitoring data sets to generate a consistent clustering network layer; If the authentication fails, a fine-tuning scale is determined based on the consistency authentication result, and the first extraction result is iteratively adjusted according to the fine-tuning scale until the authentication passes, thereby generating the consistency clustering network layer.
5. The intelligent early warning method for wind turbines according to claim 1, characterized in that: include: Matching the K clustered key component monitoring data sets with the multiple wind turbine layout locations to obtain K wind turbine layout location sets; Determining K transmission distance sets based on the K wind turbine layout position sets and the layout position of the data processing unit; Counting the number of wind turbines in the K divided wind turbine sets respectively to obtain the number of K wind turbines; Performing processing resource configuration based on the K transmission distance sets and the number of the K wind turbines to obtain K processing allocation resources; The K data processing components are constructed based on the K processing allocation resources.
6. The intelligent early warning method for wind turbines according to claim 1, characterized in that: include: Taking the key components as indexes, respectively performing the same key component monitoring data fluctuation analysis on the K clustered key component monitoring data sets, and determining K same key component fluctuation factor sets; Performing key component monitoring frequency configuration based on the K key component fluctuation factor sets to generate K component monitoring frequency sets; Asynchronous monitoring configuration is performed according to the K component monitoring frequency sets to generate asynchronous monitoring solutions for the K key components.
7. The intelligent early warning method for a wind turbine according to claim 6, characterized in that: include: Extracting K first component monitoring frequencies from the K component monitoring frequency sets respectively; Taking the K first component monitoring frequencies as K comparison objects, and taking the component monitoring frequencies in the K component monitoring frequency sets whose similarities with the K comparison objects are outside a preset similarity threshold as K first initial synchronization component monitoring frequency sets; Combining the K first key components corresponding to the K first component monitoring frequencies and the K first initial synchronization key components corresponding to the K first initial synchronization component monitoring frequency sets as K first initial synchronization monitoring components; Eliminate the component monitoring frequencies corresponding to the K first initial synchronous monitoring component combinations from the K component monitoring frequency sets to obtain K sequential monitoring frequency sets; Sorting the K sequential key component sets corresponding to the K sequential monitoring frequency sets in ascending order of monitoring frequency to obtain K first-order monitoring component sequences; The K key component asynchronous monitoring schemes are generated according to the K first initial synchronous monitoring component combinations and the K first sequential monitoring component sequences.
8. An intelligent early warning system for a wind turbine, characterized in that: The system is communicatively connected with a data processing unit, and the system comprises: A layout information acquisition module: obtains cluster layout information of a target wind turbine cluster, wherein the cluster layout information includes a plurality of wind turbines and a plurality of wind turbine layout positions; A monitoring value extraction module: extracting monitoring values of a plurality of key component sets of the plurality of wind turbines respectively according to a preset component monitoring index set at a current moment, and generating a plurality of key component monitoring data sets; A monitoring cluster division module: performs consistency clustering authentication on the plurality of key component monitoring data sets, and divides the plurality of wind turbines into monitoring clusters according to the authentication results to obtain K monitoring clusters, wherein the K monitoring clusters include K clustered key component monitoring data sets and K divided wind turbine sets; A data processing unit configuration module: configuring K data processing components of the data processing unit according to the K clustered key component monitoring data sets, the K divided wind turbine sets and the plurality of wind turbine layout positions, wherein the K data processing components correspond one-to-one to the K monitoring clusters; A key component asynchronous monitoring configuration module: traverses the K clustered key component monitoring data sets to perform key component asynchronous monitoring configuration, and obtains K key component asynchronous monitoring schemes of K monitoring clusters; Early warning analysis module: based on the K key component asynchronous monitoring schemes, key component monitoring data of the K divided wind turbine sets are collected to obtain K key component asynchronous monitoring data sets, and the K key component asynchronous monitoring data sets are input into the K data processing components for early warning analysis to obtain multiple monitoring early warning results.
