Optimized operation control method for wind turbine generator

Abstract

The invention provides a wind turbine generator optimized operation control method, and belongs to the technical field of optimized control, and the method comprises the steps: obtaining the real-time operation data of a wind turbine generator in a specified time period, and obtaining the real-time environment data of the wind turbine generator in the specified time period; determining a real-time operation vector and a real-time environment vector of the wind turbine generator; extracting historical operation reports of the wind turbine generator in a plurality of specified time periods, and determining a historical environment matrix; determining an optimal operation vector of each environment category vector in the historical environment matrix based on the multiple pieces of historical operation data; and performing optimization control on the wind turbine generator, and generating a real-time operation report. The environment change can be dynamically responded, the control strategy can be determined, the real-time operation of the wind turbine generator can be adaptively optimized and close to the optimal historical performance, the wind energy utilization efficiency is improved, the service life of equipment is prolonged, an intelligent and efficient control mode is achieved, energy consumption is reduced, and the service life of the equipment is prolonged.

Description

Technical Field

[0001] The present invention relates to the technical field of optimization control, and in particular to a method for optimizing operation control of a wind turbine generator set. Background Art

[0002] In the field of wind turbine operation control, traditional technologies usually rely on fixed control strategies or simple feedback adjustments, and lack the ability to respond to changes in the external environment in real time. In traditional methods, wind turbines are mainly set to fixed power output or speed control based on current environmental parameters such as wind speed and temperature, and the correlation between historical operation data and environmental data is not fully utilized. In actual applications, it is difficult for the unit to adapt to complex and changing environments, resulting in low operating efficiency or increased equipment wear.

[0003] Therefore, the present invention provides a method for controlling optimized operation of a wind turbine generator set. Summary of the invention

[0004] The present invention provides a method for optimizing the operation and control of a wind turbine generator set. The method determines a real-time operation vector and a real-time environment vector of the wind turbine generator set, determines a historical environment matrix of multiple historical operation reports, and determines an optimal operation vector for each environment category vector in the historical environment matrix. Based on the real-time environment vector, the real-time operation vector and the optimal operation vector of all historical environment vectors in the historical environment matrix, the wind turbine generator set is optimized and controlled, and a real-time operation report is generated. The method can dynamically respond to environmental changes and determine a control strategy so that the real-time operation of the wind turbine generator set can be adaptively optimized and close to the optimal historical performance, thereby improving wind energy utilization efficiency and extending equipment life, realizing an intelligent and efficient control mode, reducing energy consumption and extending equipment life.

[0005] The present invention provides a method for optimizing the operation and control of a wind turbine generator set, comprising:

[0006] 101: Acquire real-time operation data of the wind turbine generator set within a specified time period, and at the same time, acquire real-time environmental data of the wind turbine generator set within the specified time period;

[0007] 102: Determine a real-time operation vector of the wind turbine generator set based on the real-time operation data, and determine a real-time environment vector of the wind turbine generator set based on the real-time environment data;

[0008] 103: extracting historical operation reports of the wind turbine generator set within a plurality of specified time periods, and determining a historical environment matrix based on the plurality of historical operation reports;

[0009] 104: Determine an optimal operation vector for each environment category vector in the historical environment matrix based on a plurality of historical operation data;

[0010] 105: Based on the real-time environment vector, the real-time operation vector and the best operation vector of all historical environment vectors in the historical environment matrix, the wind turbine is optimized and controlled, and a real-time operation report is generated.

[0011] According to a wind turbine optimization operation control method provided by the present invention, real-time operation data of the wind turbine is obtained within a specified time period, and at the same time, real-time environmental data of the wind turbine is obtained within the specified time period, including:

[0012] Analyze the component structure and component characteristics of the wind turbine generator set, determine a first installation array of a first sensor group for collecting operation data based on the analysis results, and determine a key position set for collecting environmental data and a second sub-sensor group for each key position in the key position set based on the analysis results;

[0013] All sensors in the first sensor group are installed based on the first installation array respectively, and at the same time, all sensors in the corresponding second sub-sensor group are installed for each key position in the key position set respectively;

[0014] Based on the installed first sensor group, real-time sub-operation data of each component in the wind turbine generator set within a specified time period is collected. At the same time, based on the installed second sensor group, real-time sub-environment data of each key position of the wind turbine generator set within a specified time period is collected.

[0015] The real-time operation data is determined based on the real-time sub-operation data of all components, and at the same time, the real-time environment data is determined based on the real-time sub-environment data of all key locations in the key location set.

[0016] According to a wind turbine optimization operation control method provided by the present invention, a real-time operation vector of the wind turbine is determined based on real-time operation data, and a real-time environment vector of the wind turbine is determined based on real-time environment data, including:

[0017] Performing feature extraction on the real-time sub-operation data of each component in the preprocessed real-time operation data, determining a first specified number of feature values ​​of each real-time sub-operation data, and determining a sub-operation vector of each real-time sub-operation data based on the first specified number of feature values ​​of each real-time sub-operation data;

[0018] Determine the first priority sequence of all components in the wind turbine based on the importance of each component in the wind turbine to the operation of the wind turbine;

[0019] splicing the sub-operation vectors of the real-time sub-operation data of all components based on the first priority sequences of all components, and determining the real-time operation vector of the real-time operation data based on the splicing result;

[0020] Performing feature extraction on the real-time sub-environment data of each key position in the preprocessed real-time environment data, determining a second specified number of feature values ​​of each real-time sub-environment data, and determining a sub-environment vector of each real-time sub-environment data based on the second specified number of feature values ​​of each real-time sub-environment data;

[0021] Determine a second priority sequence of all key positions in the wind turbine generator set based on the importance of each key position in the key position set to the operation of the wind turbine generator set;

[0022] Based on the second priority sequence of all key positions, the sub-environment vectors of the real-time sub-environment data of all key positions are spliced, and the real-time environment vector of the real-time environment data is determined based on the splicing result.

[0023] According to a wind turbine optimization operation control method provided by the present invention, historical operation reports of the wind turbine in multiple specified time periods are extracted, and a historical environment matrix is ​​determined based on the multiple historical operation reports, including:

[0024] Extract historical operation reports of wind turbines in multiple specified time periods between the current specified time period;

[0025] Extract all historical environment vectors from multiple historical operation reports, determine the similarity between every two historical environment vectors, and sort all similarities from largest to smallest;

[0026] Based on the sorting results, two historical environment vectors corresponding to similarity are merged into a cluster family one by one, and the historical environment vectors before merging are marked. After all historical environment vectors are marked, the pre-clustering results are determined based on all the merged cluster families and the number of merged cluster families.

[0027] Comparing the number of clusters in each merged pre-clustering result with the first preset number of clusters and the second preset number of clusters;

[0028] If the number of clusters is greater than the preset number of clusters, iterative clustering is performed on the pre-clustering results corresponding to the number of clusters less than the first preset number of clusters;

[0029] If the number of clusters is less than the first preset number of clusters and greater than the second preset number of clusters, determining the pre-clustering result corresponding to the number of clusters whose number of clusters is less than the first preset number of clusters and greater than the second preset number of clusters as the actual clustering result;

[0030] If the number of clusters is less than the first preset number of clusters and less than or equal to the second preset number of clusters, determining the pre-clustering result of the last iteration of the pre-clustering result corresponding to the number of clusters less than the first preset number of clusters and less than the second preset number of clusters as the actual clustering result;

[0031] Determine the historical environment matrix based on the actual clustering results.

[0032] According to a wind turbine optimized operation control method provided by the present invention, if the number of clusters is less than a preset number of clusters, iterative clustering is performed on the pre-clustering results corresponding to the number of clusters less than the first preset number of clusters, including:

[0033] Extracting the similarity of any two historical environment vectors from all the historical environment vectors contained in every two clusters in the pre-clustering result, and adding all the extracted similarities to determine the similarity and of every two clusters in the pre-clustering result;

[0034] All similarities and are sorted from large to small. Based on the sorting results, the two clusters corresponding to the similarities and are merged into one cluster one by one, and the clusters before the merger are marked. After all the clusters before the merger are marked, the pre-clustering results of iterative clustering are determined based on all the clusters after the merger and the number of clusters after the merger.

[0035] According to a wind turbine optimization operation control method provided by the present invention, a historical environment matrix is ​​determined based on actual clustering results, including:

[0036] Determine the actual clustering result to determine the environment category, wherein each cluster family in the actual clustering result corresponds to an environment subcategory in the environment category;

[0037] Based on all historical environmental vectors contained in each cluster group in the actual clustering results, determine the environmental category vector of the environmental subcategory corresponding to each cluster group;

[0038] fc i ={fc i1 ,…,fc ij ,…,fc iN2};

[0039] Among them, fc i The environmental category vector representing the environmental subcategory corresponding to the i-th cluster family in the actual clustering result, fc i1 、fc ij 、fc iN2 They represent the fitted eigenvalues ​​of the 1st feature, the jth feature, and the N2th feature in the environmental category vector of the environmental subcategory corresponding to the i-th cluster family, respectively. N2 represents the number of features in the environmental category vector, and fc1 ij 、fc2 ij They represent the first eigenvalue and the second eigenvalue of the jth feature in the environment category vector of the environment subcategory corresponding to the i-th cluster family, w1 and w2 represent the first weight of the first eigenvalue and the second weight of the second eigenvalue, respectively. ijkrepresents the average feature value of the jth feature of the kth historical environmental vector in the environmental subcategory corresponding to the ith cluster family, iN1 represents the number of all historical environmental vectors in the environmental subcategory corresponding to the ith cluster family, V iko represents the similarity between the kth historical environment vector and the oth historical environment vector in the i-th corresponding environment subcategory, α represents the similarity conversion parameter, represents the average eigenvalue of the jth feature of all historical environmental vectors in the environmental subcategory corresponding to the i-th cluster family, σf ijk represents the standard deviation of the jth feature of all historical environmental vectors in the environmental subcategory corresponding to the i-th cluster family, γ represents the adjustment parameter of the standard deviation, med(f ijk )、max(f ijk )、min(f ijk ) represent the median, maximum, and minimum values ​​of the jth feature of all historical environmental vectors in the environmental subcategory corresponding to the i-th cluster family, τ represents the power adjustment parameter, and μ represents the truncation adjustment parameter;

[0040] A historical environment matrix for multiple historical run reports is determined based on the environmental category vector results of all cluster families in the actual clustering results.

[0041] According to a wind turbine optimization operation control method provided by the present invention, the optimal operation vector of each historical environment vector in the historical environment matrix is ​​determined based on a plurality of historical operation data, including:

[0042] Extract the historical operation vector corresponding to each historical environment vector contained in each cluster family in the historical environment matrix;

[0043] Determine the historical operating state corresponding to each historical operating vector of each cluster family extracted, sort all historical operating states of each cluster family, and determine that the historical operating vector corresponding to the sorted third quartile is the best operating vector of the environmental category vector of the environmental subcategory corresponding to each cluster family.

[0044] According to a wind turbine optimization operation control method provided by the present invention, based on the real-time environment vector, the real-time operation vector and the best operation vector of all historical environment vectors in the historical environment matrix, the wind turbine is optimized and controlled, and a real-time operation report is generated, including:

[0045] Matching the real-time environment vector with the environment category vectors of all environment subcategories in the historical environment matrix, and determining the environment category vector with the highest matching degree as the first environment vector of the real-time environment vector;

[0046] Determine the best operation vector of the environment category vector corresponding to the first environment vector as the first operation vector of the real-time operation vector;

[0047] Determine an optimal control strategy for the wind turbine generator set based on the real-time operation vector and the first operation vector, and perform optimal control on the wind turbine generator set based on the control strategy;

[0048] A real-time operation report is generated based on the real-time operation data, the real-time operation vector, the real-time environment data, the real-time environment vector, the first environment vector, the first operation vector, and the control strategy.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] By determining the real-time operation vector and real-time environment vector of the wind turbine, determining the historical environment matrix of multiple historical operation reports, and determining the optimal operation vector of each environment category vector in the historical environment matrix, the wind turbine is optimized and controlled based on the real-time environment vector, the real-time operation vector and the optimal operation vector of all historical environment vectors in the historical environment matrix, and a real-time operation report is generated. This can dynamically respond to environmental changes and determine control strategies so that the real-time operation of the wind turbine can be adaptively optimized and close to the optimal historical performance, thereby improving wind energy utilization efficiency and extending equipment life, realizing an intelligent and efficient control mode, reducing energy consumption and extending equipment life. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0052] Figure 1 It is a flow chart of a method for optimizing operation and controlling a wind turbine set provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] Embodiment 1:

[0055] The embodiment of the present invention provides a method for optimizing the operation control of a wind turbine generator system. Figure 1 As shown, including:

[0056] 101: Acquire real-time operation data of the wind turbine generator set within a specified time period, and at the same time, acquire real-time environmental data of the wind turbine generator set within the specified time period;

[0057] 102: Determine a real-time operation vector of the wind turbine generator set based on the real-time operation data, and determine a real-time environment vector of the wind turbine generator set based on the real-time environment data;

[0058] 103: extracting historical operation reports of the wind turbine generator set within a plurality of specified time periods, and determining a historical environment matrix based on the plurality of historical operation reports;

[0059] 104: Determine an optimal operation vector for each environment category vector in the historical environment matrix based on a plurality of historical operation data;

[0060] 105: Based on the real-time environment vector, the real-time operation vector and the best operation vector of all historical environment vectors in the historical environment matrix, the wind turbine is optimized and controlled, and a real-time operation report is generated.

[0061] In this embodiment, for each environment category vector in the historical environment matrix, a corresponding optimal operating vector is determined based on its performance under different historical operating states. This vector represents the optimal historical operating state of the wind turbine under this environment category, providing a reference for optimized control.

[0062] In this embodiment, each environment category vector in the historical environment matrix corresponds to an optimal operation vector.

[0063] In this embodiment, an optimization control strategy is determined by comparing the real-time environment vector, the real-time operation vector and the optimal operation vector of each environment category in the historical environment matrix. This strategy makes the real-time operation status of the wind turbine as close as possible to the historical optimal performance, and finally generates a real-time operation report summarizing the current operation status, optimization measures and execution effects.

[0064] The beneficial effects of the above technical solution are as follows: by determining the real-time operation vector and real-time environment vector of the wind turbine, determining the historical environment matrix of multiple historical operation reports, and determining the optimal operation vector of each environment category vector in the historical environment matrix, based on the real-time environment vector, the real-time operation vector and the optimal operation vector of all historical environment vectors in the historical environment matrix, the wind turbine is optimized and controlled, and a real-time operation report is generated. It can dynamically respond to environmental changes and determine control strategies so that the real-time operation of the wind turbine can be adaptively optimized and close to the best historical performance, thereby improving wind energy utilization efficiency and extending equipment life, realizing an intelligent and efficient control mode, reducing energy consumption and extending equipment life.

[0065] Embodiment 2:

[0066] The embodiment of the present invention provides a method for optimizing the operation and control of a wind turbine generator set, which obtains real-time operation data of the wind turbine generator set within a specified time period, and at the same time, obtains real-time environmental data of the wind turbine generator set within the specified time period, including:

[0067] Analyze the component structure and component characteristics of the wind turbine generator set, determine a first installation array of a first sensor group for collecting operation data based on the analysis results, and determine a key position set for collecting environmental data and a second sub-sensor group for each key position in the key position set based on the analysis results;

[0068] All sensors in the first sensor group are installed based on the first installation array respectively, and at the same time, all sensors in the corresponding second sub-sensor group are installed for each key position in the key position set respectively;

[0069] Based on the installed first sensor group, real-time sub-operation data of each component in the wind turbine generator set within a specified time period is collected. At the same time, based on the installed second sensor group, real-time sub-environment data of each key position of the wind turbine generator set within a specified time period is collected.

[0070] The real-time operation data is determined based on the real-time sub-operation data of all components, and at the same time, the real-time environment data is determined based on the real-time sub-environment data of all key locations in the key location set.

[0071] In this embodiment, the first sensor group is a sensor set used to collect operating data of various components of the wind turbine generator set.

[0072] In this embodiment, the first installation array represents a specific position array where the first sensor group is installed to ensure data collection of each component.

[0073] In this embodiment, the key position refers to a special position inside or outside the wind turbine generator set, and each key position may have a significant impact on the performance of the wind turbine generator set.

[0074] In this embodiment, the second sub-sensor group refers to an environmental sensor group installed at a key position, which is specifically used to collect environmental data.

[0075] In this embodiment, the structure and characteristics of different components of the wind turbine (such as blades, generators, pitch control systems, etc.) are first analyzed to determine the type of data that needs to be collected for each component (such as temperature, vibration, rotation speed, etc.).

[0076] In this embodiment, based on the component analysis results, the first sensor group required to collect the operating data of all components is determined, and a first installation array suitable for the data collection points of each component is designed. A suitable data sensor is placed at each designated installation point to achieve accurate monitoring of the component.

[0077] In this embodiment, in the surrounding or internal environment of the wind turbine, key locations (such as air inlet, air outlet, cooling system, etc.) are selected based on structural and environmental analysis. The environment of each key location may be different, so a suitable second sub-sensor group is arranged at each key location to collect environmental data (such as temperature, wind speed, humidity, etc.).

[0078] The beneficial effects of the above technical solution are: obtaining real-time operation data of the wind turbine set within a specified time period, and at the same time, obtaining real-time environmental data of the wind turbine set within a specified time period, which can accurately and comprehensively collect wind point unit data and provide data support for optimized control.

[0079] Embodiment 3:

[0080] An embodiment of the present invention provides a method for optimizing operation and control of a wind turbine generator set, which determines a real-time operation vector of the wind turbine generator set based on real-time operation data, and determines a real-time environment vector of the wind turbine generator set based on real-time environment data, including:

[0081] Performing feature extraction on the real-time sub-operation data of each component in the preprocessed real-time operation data, determining a first specified number of feature values ​​of each real-time sub-operation data, and determining a sub-operation vector of each real-time sub-operation data based on the first specified number of feature values ​​of each real-time sub-operation data;

[0082] Determine the first priority sequence of all components in the wind turbine based on the importance of each component in the wind turbine to the operation of the wind turbine;

[0083] splicing the sub-operation vectors of the real-time sub-operation data of all components based on the first priority sequences of all components, and determining the real-time operation vector of the real-time operation data based on the splicing result;

[0084] Performing feature extraction on the real-time sub-environment data of each key position in the preprocessed real-time environment data, determining a second specified number of feature values ​​of each real-time sub-environment data, and determining a sub-environment vector of each real-time sub-environment data based on the second specified number of feature values ​​of each real-time sub-environment data;

[0085] Determine a second priority sequence of all key positions in the wind turbine generator set based on the importance of each key position in the key position set to the operation of the wind turbine generator set;

[0086] Based on the second priority sequence of all key positions, the sub-environment vectors of the real-time sub-environment data of all key positions are spliced, and the real-time environment vector of the real-time environment data is determined based on the splicing result.

[0087] In this embodiment, feature extraction is performed on the preprocessed real-time operation data, and a first specified number of characteristic values ​​(such as the mean temperature, frequency characteristics of vibration, etc.) are extracted from the real-time sub-operation data of each component. These characteristic values ​​reflect the main operating status of the component within a specified time period. The extracted characteristic values ​​are used to generate a sub-operation vector for each component to describe the real-time operating status of each component.

[0088] In this embodiment, the first priority sequence of each component is set according to the influence of each component on the overall performance during the operation of the wind turbine generator set, that is, the components are sorted according to their importance.

[0089] In this embodiment, the sub-operation vectors of each component are spliced ​​in sequence according to the first priority sequence to form a complete real-time operation vector, which integrates the operation status of all components. The greater the impact of the component on the overall performance, the closer the sub-operation vector corresponding to the component is to the front of the real-time operation vector.

[0090] In this embodiment, feature extraction is performed on the real-time sub-environmental data of each key location of the real-time environmental data, and a second specified number of feature values ​​(such as the mean of wind speed, humidity fluctuation, etc.) are extracted to reflect the environmental status of the key location.

[0091] In this embodiment, a sub-environment vector of each key location is generated based on the extracted feature values ​​to describe the environmental conditions of the location.

[0092] In this embodiment, a second priority sequence is generated according to the degree of influence of each key position on the overall operation of the wind turbine generator system.

[0093] In this embodiment, according to the second priority sequence, the sub-environment vectors of all key positions are sequentially spliced ​​to form a complete real-time environment vector, and the greater the importance of the key position, the closer the sub-environment vector of the key position is to the front of the real-time environment vector.

[0094] The beneficial effects of the above technical solution are: determining the real-time operation vector of the wind turbine based on the real-time operation data, and determining the real-time environmental vector of the wind turbine based on the real-time environmental data, achieving efficient data processing and accurate status expression, improving the real-time control of the operating status of the wind turbine, and providing accurate data support for optimized control.

[0095] Embodiment 4:

[0096] An embodiment of the present invention provides a method for optimizing operation and control of a wind turbine generator set, extracting historical operation reports of the wind turbine generator set within multiple specified time periods, and determining a historical environment matrix based on the multiple historical operation reports, including:

[0097] Extract historical operation reports of wind turbines in multiple specified time periods between the current specified time period;

[0098] Extract all historical environment vectors from multiple historical operation reports, determine the similarity between every two historical environment vectors, and sort all similarities from largest to smallest;

[0099] Based on the sorting results, two historical environment vectors corresponding to similarity are merged into a cluster family one by one, and the historical environment vectors before merging are marked. After all historical environment vectors are marked, the pre-clustering results are determined based on all the merged cluster families and the number of merged cluster families.

[0100] Comparing the number of clusters in each merged pre-clustering result with the first preset number of clusters and the second preset number of clusters;

[0101] If the number of clusters is greater than the preset number of clusters, iterative clustering is performed on the pre-clustering results corresponding to the number of clusters less than the first preset number of clusters;

[0102] If the number of clusters is less than the first preset number of clusters and greater than the second preset number of clusters, determining the pre-clustering result corresponding to the number of clusters whose number of clusters is less than the first preset number of clusters and greater than the second preset number of clusters as the actual clustering result;

[0103] If the number of clusters is less than the first preset number of clusters and less than or equal to the second preset number of clusters, determining the pre-clustering result of the last iteration of the pre-clustering result corresponding to the number of clusters less than the first preset number of clusters and less than the second preset number of clusters as the actual clustering result;

[0104] Determine the historical environment matrix based on the actual clustering results.

[0105] In this embodiment, if the number of clusters is less than the first preset number of clusters and less than or equal to the second preset number of clusters, the pre-clustering result of the previous iteration corresponding to the pre-clustering result in which the number of clusters is less than the first preset number of clusters and less than or equal to the second preset number of clusters is determined to be the actual clustering result. For example: the first preset number of clusters is 10, the second preset number of clusters is 7, the number of clusters in a certain pre-clustering result is 12, 12>10, so iterative clustering is performed again, and the number of clusters in the pre-clustering result of the iterative clustering result is 6, 6<12 and 6<7, so it is determined that the number of 12 clusters in the pre-clustering result of the previous iteration in which the iterative clustering result is performed again and the number of clusters 12 are the actual clustering results.

[0106] In this embodiment, historical environment vectors (each vector represents the environment status of a time period) are extracted from all historical operation reports, the similarity between every two historical environment vectors is calculated (such as using a distance or similarity formula to calculate the similarity value), and the similarity values ​​are sorted from large to small.

[0107] In this embodiment, two historical environment vectors with high similarity are gradually merged into a cluster by sorting by similarity, and the merged vectors are marked. This process continues until all historical environment vectors are marked.

[0108] In this embodiment, the pre-clustering result consists of the clusters formed by merging and the number of clusters.

[0109] In this embodiment, the number of clusters in the current pre-clustering result is compared with the first preset number of clusters and the second preset number of clusters:

[0110] When the number of clusters is greater than the preset value, the clusters in the current result are iteratively clustered and further subdivided;

[0111] When the number of clusters is less than the first preset number of clusters and less than the second preset number of clusters, clustering is stopped and the current pre-clustering result is recognized as the actual clustering result.

[0112] When the number of cluster groups is less than the first preset number of cluster groups and less than the second preset number of cluster groups, the pre-clustering result of the previous iteration is rolled back and used as the final actual clustering result.

[0113] The beneficial effects of the above technical solution are as follows: extracting historical operation reports of wind turbines within multiple specified time periods and determining the historical environment matrix based on multiple historical operation reports can improve the accuracy of identifying the environmental status of wind turbines and enhance the practicality and accuracy of historical environmental data analysis.

[0114] Embodiment 5:

[0115] An embodiment of the present invention provides a method for controlling optimized operation of a wind turbine generator set. If the number of clusters is less than a preset number of clusters, iterative clustering is performed on pre-clustering results corresponding to the number of clusters less than a first preset number of clusters, including:

[0116] Extracting the similarity of any two historical environment vectors from all the historical environment vectors contained in every two clusters in the pre-clustering result, and adding all the extracted similarities to determine the similarity and of every two clusters in the pre-clustering result;

[0117] All similarities and are sorted from large to small. Based on the sorting results, the two clusters corresponding to the similarities and are merged into one cluster one by one, and the clusters before the merger are marked. After all the clusters before the merger are marked, the pre-clustering results of iterative clustering are determined based on all the clusters after the merger and the number of clusters after the merger.

[0118] In this embodiment, for each cluster in the pre-clustering result, for every two clusters, the similarity between all historical environment vectors contained in them is extracted, and each pair of similarity results is added to obtain the similarity and between the two clusters, which represents how close the two clusters are overall.

[0119] In this embodiment, all clusters are sorted from large to small in similarity, so that clusters with higher similarity are merged first, and two clusters with higher similarity are gradually merged according to the sorting results, and the merged clusters are marked to avoid repeated processing.

[0120] In this embodiment, merging is continued until all pre-merged clusters are marked, indicating that all possible cluster merging has been completed. Based on the merged clusters and their numbers, the pre-clustering results of the iterative clustering obtained in this round of clustering are determined. The results include the merged new clusters for further analysis.

[0121] In this embodiment, for example, the pre-clustering result includes 12 clusters, the similarities of all historical environment vectors included in every two clusters of the 12 clusters are extracted, all similarities of the extracted two clusters are added, and the similarities and of the two clusters are determined.

[0122] The beneficial effect of the above technical solution is: if the number of clusters is less than the preset number of clusters, iterative clustering is performed on the pre-clustering results corresponding to the number of clusters less than the first preset number of clusters, which can enhance the clustering accuracy of environmental data and make the analysis results of historical environmental status more representative and stable.

[0123] Embodiment 6:

[0124] An embodiment of the present invention provides a method for optimizing operation and control of a wind turbine generator set, which determines a historical environment matrix based on actual clustering results, and includes:

[0125] Determine the actual clustering result to determine the environment category, wherein each cluster family in the actual clustering result corresponds to an environment subcategory in the environment category;

[0126] Based on all historical environmental vectors contained in each cluster group in the actual clustering results, determine the environmental category vector of the environmental subcategory corresponding to each cluster group;

[0127] fc i ={fc i1 ,…,fc ij,…,fc iN2};

[0128] Among them, fc i The environmental category vector representing the environmental subcategory corresponding to the i-th cluster family in the actual clustering result, fc i1 、fc ij 、fc iN2 They represent the fitted eigenvalues ​​of the 1st feature, the jth feature, and the N2th feature in the environmental category vector of the environmental subcategory corresponding to the i-th cluster family, respectively. N2 represents the number of features in the environmental category vector, and fc1 ij 、fc2 ij They represent the first eigenvalue and the second eigenvalue of the jth feature in the environment category vector of the environment subcategory corresponding to the i-th cluster family, w1 and w2 represent the first weight of the first eigenvalue and the second weight of the second eigenvalue, respectively. ijk represents the average feature value of the jth feature of the kth historical environmental vector in the environmental subcategory corresponding to the ith cluster family, iN1 represents the number of all historical environmental vectors in the environmental subcategory corresponding to the ith cluster family, V iko represents the similarity between the kth historical environment vector and the oth historical environment vector in the i-th corresponding environment subcategory, α represents the similarity conversion parameter, represents the average eigenvalue of the jth feature of all historical environmental vectors in the environmental subcategory corresponding to the i-th cluster family, σf ijk represents the standard deviation of the jth feature of all historical environmental vectors in the environmental subcategory corresponding to the i-th cluster family, γ represents the adjustment parameter of the standard deviation, med(f ijk )、max(f ijk )、min(f ijk ) represent the median, maximum, and minimum values ​​of the jth feature of all historical environmental vectors in the environmental subcategory corresponding to the i-th cluster family, τ represents the power adjustment parameter, and μ represents the truncation adjustment parameter;

[0129] A historical environment matrix for multiple historical run reports is determined based on the environmental category vector results of all cluster families in the actual clustering results.

[0130] In this embodiment, each cluster group in the actual clustering result corresponds to an environment category vector of an environment subcategory in the environment category.

[0131] In this embodiment, for each cluster group in the actual clustering result, all historical environment vectors contained in the cluster group and the similarity between every two historical environment vectors are extracted, and the environment category vector of the cluster group is calculated based on these.

[0132] In this embodiment, It represents the average value of the first μ and last μ feature values ​​of the j-th feature of all historical environment vectors in the environment subcategory corresponding to the sorted i-th cluster family after truncating them.

[0133] In this embodiment, the environment category vectors of all clusters in the actual clustering result are combined and arranged into a matrix, namely, a historical environment matrix, which is used to summarize the environment data of multiple historical operation reports.

[0134] The beneficial effects of the above technical solution are as follows: the historical environment matrix is ​​determined based on the actual clustering results, so that the analysis of the historical environment status is more representative and stable, and the recognition accuracy of the environmental status of the wind turbine is improved.

[0135] Embodiment 7:

[0136] An embodiment of the present invention provides a method for optimizing operation and control of a wind turbine generator set, which determines the best operation vector of each historical environment vector in a historical environment matrix based on a plurality of historical operation data, and includes:

[0137] Extract the historical operation vector corresponding to each historical environment vector contained in each cluster family in the historical environment matrix;

[0138] Determine the historical operating state corresponding to each historical operating vector of each cluster family extracted, sort all historical operating states of each cluster family, and determine that the historical operating vector corresponding to the sorted third quartile is the best operating vector of the environmental category vector of the environmental subcategory corresponding to each cluster family.

[0139] In this embodiment, in the historical environment matrix, each cluster contains a plurality of historical environment vectors, and for these environment vectors, corresponding historical operation vectors are extracted one by one, ie, the operation status data of the wind turbine generator set under the environment state.

[0140] In this embodiment, the extracted historical operating states are sorted according to performance (such as key indicators such as power and efficiency), and the sorting method is from high to low.

[0141] In this embodiment, for the historical operating status of each cluster group after sorting, the historical operating vector corresponding to the third quartile is taken as the optimal operating vector of the cluster group.

[0142] In this embodiment, the best operation vector is a representative vector of the operation performance of the environment subcategory, and generally has a good performance level, indicating a better operation state in the environment.

[0143] The beneficial effects of the above technical solution are as follows: the optimal operating vector of each historical environment vector in the historical environment matrix is ​​determined based on multiple historical operating data, which provides a scientific basis for determining the optimal control strategy of the wind turbine and helps to more accurately match the optimal operating parameters of different environments.

[0144] Embodiment 8:

[0145] The embodiment of the present invention provides a method for optimizing the operation and control of a wind turbine generator set, which optimizes and controls the wind turbine generator set based on a real-time environment vector, a real-time operation vector, and an optimal operation vector of all historical environment vectors in a historical environment matrix, and generates a real-time operation report, including:

[0146] Matching the real-time environment vector with the environment category vectors of all environment subcategories in the historical environment matrix, and determining the environment category vector with the highest matching degree as the first environment vector of the real-time environment vector;

[0147] Determine the best operation vector of the environment category vector corresponding to the first environment vector as the first operation vector of the real-time operation vector;

[0148] Determine an optimal control strategy for the wind turbine generator set based on the real-time operation vector and the first operation vector, and perform optimal control on the wind turbine generator set based on the control strategy;

[0149] A real-time operation report is generated based on the real-time operation data, the real-time operation vector, the real-time environment data, the real-time environment vector, the first environment vector, the first operation vector, and the control strategy.

[0150] In this embodiment, the real-time environment vector is a state description of the wind turbine in the current real-time environment. By matching the environment category vectors of all environment subcategories in the historical environment matrix, the similarity between each category vector and the real-time environment vector is calculated, and the environment category vector with the highest matching degree is found, which is called the first environment vector, which represents the historical environment state closest to the current environment.

[0151] In this embodiment, after finding the first environmental vector, the corresponding optimal operating vector is determined, that is, the historical operating vector with the best performance of the wind turbine under this environmental category. This vector is used as the first operating vector of the real-time environment. The first operating vector represents the best historical operating performance of the unit under similar environments and provides a reference for current real-time control.

[0152] In this embodiment, based on the comparison between the real-time operation vector and the first operation vector, it is determined whether the current unit control strategy needs to be adjusted.

[0153] In this embodiment, if the real-time operating performance is lower than the historical best performance, the optimization control strategy will give adjustment suggestions, such as adjusting the speed, power output, etc., to make the real-time performance as close to the optimal level as possible.

[0154] The beneficial effects of the above technical solution are as follows: based on the real-time environment vector, the real-time operation vector and the optimal operation vector of all historical environment vectors in the historical environment matrix, the wind turbine is optimized and controlled, and a real-time operation report is generated. It can dynamically respond to environmental changes and determine the control strategy so that the real-time operation of the wind turbine can be adaptively optimized and close to the best historical performance, thereby improving the efficiency of wind energy utilization and extending the life of the equipment, realizing an intelligent and efficient control mode, reducing energy consumption and extending the life of the equipment.

[0155] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0156] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0157] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing the operation and control of a wind turbine generator system, characterized in that: include: 101: Acquire real-time operation data of the wind turbine generator set within a specified time period, and at the same time, acquire real-time environmental data of the wind turbine generator set within the specified time period; 102: Determine a real-time operation vector of the wind turbine generator set based on the real-time operation data, and determine a real-time environment vector of the wind turbine generator set based on the real-time environment data; 103: extracting historical operation reports of the wind turbine generator set within a plurality of specified time periods, and determining a historical environment matrix based on the plurality of historical operation reports; 104: Determine an optimal operation vector for each environment category vector in the historical environment matrix based on a plurality of historical operation data; 105: Based on the real-time environment vector, the real-time operation vector and the best operation vector of all historical environment vectors in the historical environment matrix, the wind turbine is optimized and controlled, and a real-time operation report is generated.

2. A wind turbine optimized operation control method according to claim 1, characterized in that: Obtain the real-time operation data of the wind turbine in a specified time period, and at the same time, obtain the real-time environmental data of the wind turbine in a specified time period, including: Analyze the component structure and component characteristics of the wind turbine generator set, determine a first installation array of a first sensor group for collecting operation data based on the analysis results, and determine a key position set for collecting environmental data and a second sub-sensor group for each key position in the key position set based on the analysis results; All sensors in the first sensor group are installed based on the first installation array respectively, and at the same time, all sensors in the corresponding second sub-sensor group are installed for each key position in the key position set respectively; Based on the installed first sensor group, real-time sub-operation data of each component in the wind turbine generator set within a specified time period is collected. At the same time, based on the installed second sensor group, real-time sub-environment data of each key position of the wind turbine generator set within a specified time period is collected. The real-time operation data is determined based on the real-time sub-operation data of all components, and at the same time, the real-time environment data is determined based on the real-time sub-environment data of all key locations in the key location set.

3. A wind turbine optimized operation control method according to claim 2, characterized in that: The real-time operation vector of the wind turbine generator set is determined based on the real-time operation data, and the real-time environment vector of the wind turbine generator set is determined based on the real-time environment data, including: Performing feature extraction on the real-time sub-operation data of each component in the preprocessed real-time operation data, determining a first specified number of feature values ​​of each real-time sub-operation data, and determining a sub-operation vector of each real-time sub-operation data based on the first specified number of feature values ​​of each real-time sub-operation data; Determine the first priority sequence of all components in the wind turbine based on the importance of each component in the wind turbine to the operation of the wind turbine; splicing the sub-operation vectors of the real-time sub-operation data of all components based on the first priority sequences of all components, and determining the real-time operation vector of the real-time operation data based on the splicing result; Performing feature extraction on the real-time sub-environment data of each key position in the preprocessed real-time environment data, determining a second specified number of feature values ​​of each real-time sub-environment data, and determining a sub-environment vector of each real-time sub-environment data based on the second specified number of feature values ​​of each real-time sub-environment data; Determine a second priority sequence of all key positions in the wind turbine generator set based on the importance of each key position in the key position set to the operation of the wind turbine generator set; Based on the second priority sequence of all key positions, the sub-environment vectors of the real-time sub-environment data of all key positions are spliced, and the real-time environment vector of the real-time environment data is determined based on the splicing result.

4. A wind turbine optimized operation control method according to claim 1, characterized in that: Extract historical operation reports of wind turbines in multiple specified time periods, and determine the historical environment matrix based on multiple historical operation reports, including: Extract historical operation reports of wind turbines in multiple specified time periods between the current specified time period; Extract all historical environment vectors from multiple historical operation reports, determine the similarity between every two historical environment vectors, and sort all similarities from largest to smallest; Based on the sorting results, two historical environment vectors corresponding to similarity are merged into a cluster family one by one, and the historical environment vectors before merging are marked. After all historical environment vectors are marked, the pre-clustering results are determined based on all the merged cluster families and the number of merged cluster families. Comparing the number of clusters in each merged pre-clustering result with the first preset number of clusters and the second preset number of clusters; If the number of clusters is greater than the preset number of clusters, iterative clustering is performed on the pre-clustering results corresponding to the number of clusters less than the first preset number of clusters; If the number of clusters is less than the first preset number of clusters and greater than the second preset number of clusters, determining the pre-clustering result corresponding to the number of clusters whose number of clusters is less than the first preset number of clusters and greater than the second preset number of clusters as the actual clustering result; If the number of clusters is less than the first preset number of clusters and less than or equal to the second preset number of clusters, determining the pre-clustering result of the last iteration of the pre-clustering result corresponding to the number of clusters less than the first preset number of clusters and less than the second preset number of clusters as the actual clustering result; Determine the historical environment matrix based on the actual clustering results.

5. A wind turbine optimized operation control method according to claim 4, characterized in that: If the number of clusters is less than the preset number of clusters, iterative clustering is performed on the pre-clustering results corresponding to the number of clusters less than the first preset number of clusters, including: Extracting the similarity of any two historical environment vectors from all the historical environment vectors contained in every two clusters in the pre-clustering result, and adding all the extracted similarities to determine the similarity and of every two clusters in the pre-clustering result; All similarities and are sorted from large to small. Based on the sorting results, the two clusters corresponding to the similarities and are merged into one cluster one by one, and the clusters before the merger are marked. After all the clusters before the merger are marked, the pre-clustering results of iterative clustering are determined based on all the clusters after the merger and the number of clusters after the merger.

6. A wind turbine optimized operation control method according to claim 4, characterized in that: Determine the historical environment matrix based on the actual clustering results, including: Determine the actual clustering result to determine the environment category, wherein each cluster family in the actual clustering result corresponds to an environment subcategory in the environment category; Based on all historical environmental vectors contained in each cluster group in the actual clustering results, determine the environmental category vector of the environmental subcategory corresponding to each cluster group; fc i ={fc i1 ,…,fc ij ,…,fc iN2 }; Among them, fc i The environmental category vector representing the environmental subcategory corresponding to the i-th cluster family in the actual clustering result, fc i1 、fc ij 、fc iN2 They represent the fitted eigenvalues ​​of the 1st feature, the jth feature, and the N2th feature in the environmental category vector of the environmental subcategory corresponding to the i-th cluster family, respectively. N2 represents the number of features in the environmental category vector, and fc1 ij 、fc2 ij They represent the first eigenvalue and the second eigenvalue of the jth feature in the environment category vector of the environment subcategory corresponding to the i-th cluster family, w1 and w2 represent the first weight of the first eigenvalue and the second weight of the second eigenvalue, respectively. ijk represents the average feature value of the jth feature of the kth historical environmental vector in the environmental subcategory corresponding to the ith cluster family, iN1 represents the number of all historical environmental vectors in the environmental subcategory corresponding to the ith cluster family, V iko represents the similarity between the kth historical environment vector and the oth historical environment vector in the i-th corresponding environment subcategory, α represents the similarity conversion parameter, represents the average eigenvalue of the jth feature of all historical environmental vectors in the environmental subcategory corresponding to the i-th cluster family, σf ijk represents the standard deviation of the jth feature of all historical environmental vectors in the environmental subcategory corresponding to the i-th cluster family, γ represents the adjustment parameter of the standard deviation, med(f ijk )、max(f ijk )、min(f ijk ) represent the median, maximum, and minimum values ​​of the jth feature of all historical environmental vectors in the environmental subcategory corresponding to the i-th cluster family, τ represents the power adjustment parameter, and μ represents the truncation adjustment parameter; A historical environment matrix for multiple historical run reports is determined based on the environmental category vector results of all cluster families in the actual clustering results.

7. A wind turbine optimized operation control method according to claim 1, characterized in that: Determine the best operation vector for each historical environment vector in the historical environment matrix based on multiple historical operation data, including: Extract the historical operation vector corresponding to each historical environment vector contained in each cluster family in the historical environment matrix; Determine the historical operating state corresponding to each historical operating vector of each cluster family extracted, sort all historical operating states of each cluster family, and determine that the historical operating vector corresponding to the sorted third quartile is the best operating vector of the environmental category vector of the environmental subcategory corresponding to each cluster family.

8. A wind turbine optimized operation control method according to claim 1, characterized in that: Based on the real-time environment vector, the real-time operation vector and the best operation vector of all historical environment vectors in the historical environment matrix, the wind turbine is optimally controlled and a real-time operation report is generated, including: Matching the real-time environment vector with the environment category vectors of all environment subcategories in the historical environment matrix, and determining the environment category vector with the highest matching degree as the first environment vector of the real-time environment vector; Determine the best operation vector of the environment category vector corresponding to the first environment vector as the first operation vector of the real-time operation vector; Determine an optimal control strategy for the wind turbine generator set based on the real-time operation vector and the first operation vector, and perform optimal control on the wind turbine generator set based on the control strategy; A real-time operation report is generated based on the real-time operation data, the real-time operation vector, the real-time environment data, the real-time environment vector, the first environment vector, the first operation vector, and the control strategy.

Images