An online intelligent monitoring method for verticality of a fan tower drum

Abstract

The present application relates to the technical field of data processing, and especially relates to an online intelligent monitoring method for the verticality of a fan tower drum, which obtains comprehensive vibration data and multi-dimensional environmental data at each monitoring time within a preset historical period and at a current monitoring time; according to the data fluctuation characteristics of the multi-dimensional environmental data and the comprehensive vibration data within the preset historical period, the influence degree of each kind of environmental data is obtained, the multi-dimensional environmental data within the preset historical period is clustered, and the normal comprehensive vibration data corresponding to each cluster is obtained; according to the difference between the multi-dimensional environmental data in each cluster and the multi-dimensional environmental data at the current monitoring time, the normal comprehensive vibration data corresponding to each cluster, and the influence degree of each kind of environmental data, the correction value corresponding to the current monitoring time is obtained; and the verticality of the fan tower drum is obtained by using the correction value corresponding to the current monitoring time of each direction at each monitoring position in the fan tower drum, thereby improving the accuracy of the tower drum verticality monitoring.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to an online intelligent monitoring method for the verticality of wind turbine towers. Background Technology

[0002] The wind turbine tower is the load-bearing structure supporting the nacelle and rotor of the wind turbine. Significant deviations in tower verticality can lead to unbalanced turbine operation, affecting power generation efficiency, and can also cause additional structural stress on the tower, impacting structural stability and even posing a risk of tower collapse. Therefore, real-time monitoring of tower verticality is crucial for improving the safety and reliability of wind turbine units.

[0003] Existing methods for monitoring the verticality of wind turbine towers mainly utilize monitoring equipment such as inclinometers and accelerometers to acquire tower vibration data in real time, and then use this vibration data to calculate the tower's verticality. However, tower vibration data is easily affected by environmental factors such as wind speed, wind direction, temperature, and humidity, causing the vibration data to not accurately reflect the true verticality of the tower.

[0004] To reduce the impact of environmental factors, existing technologies typically employ fixed threshold methods to remove some tower vibration data, or use simple linear models to compensate for environmental factors in tower vibration data. However, the combined effect of multiple environmental factors and the complexity of environmental changes make it difficult to accurately represent the impact of the environment on tower vibration data using fixed thresholds or linear models, resulting in deviations in the final tower verticality readings.

[0005] Therefore, how to reduce the impact of environmental factors on tower vibration data and improve the accuracy of tower verticality monitoring has become an urgent problem to be solved. Summary of the Invention

[0006] In view of this, embodiments of the present invention provide an online intelligent monitoring method for the verticality of wind turbine towers, in order to solve the problem of how to reduce the impact of environmental factors on tower vibration data and improve the accuracy of tower verticality monitoring.

[0007] This invention provides an online intelligent monitoring method for the verticality of wind turbine towers, which includes the following steps:

[0008] For any monitoring position in the wind turbine tower in any direction, based on the vibration data at each sampling time in any direction, obtain the comprehensive vibration data at the current monitoring time and at each monitoring time within the preset historical period, and obtain the multi-dimensional environmental data at the current monitoring time and at each monitoring time within the preset historical period.

[0009] Based on the data fluctuation characteristics of multidimensional environmental data and comprehensive vibration data within a preset historical period, the influence degree of each type of environmental data is obtained. Using the influence degree of each type of environmental data, the multidimensional environmental data within the preset historical period is clustered to obtain at least one cluster. Based on the fluctuation characteristics of the vibration data corresponding to the monitoring time of the multidimensional environmental data in each cluster, the normal comprehensive vibration data corresponding to each cluster is obtained.

[0010] Based on the difference between the multidimensional environmental data in each cluster and the multidimensional environmental data at the current monitoring time, the normal comprehensive vibration data corresponding to each cluster, and the degree of influence of each type of environmental data, the degree of influence of environmental factors at the current monitoring time is obtained. The degree of influence of environmental factors is used to correct each vibration data corresponding to the current monitoring time to obtain the correction value of each vibration data.

[0011] The verticality of the wind turbine tower is obtained by using the correction value of each vibration data corresponding to each direction at each monitoring position in the wind turbine tower at the current monitoring time.

[0012] Preferably, the step of obtaining comprehensive vibration data at the current monitoring time and at each monitoring time within a preset historical period based on the vibration data at each sampling time in any direction includes:

[0013] For the current monitoring time and any monitoring time within a preset historical period, a time window of a preset length is constructed with the any monitoring time as the cutoff time. The preset length is the interval between the any monitoring time and its previous monitoring time.

[0014] Acquire tilt monitoring data at any monitoring time. For any sampling time within the time window, use the tilt monitoring data to obtain the gravity component of the vibration data at any sampling time. Calculate the difference between the vibration data at any sampling time and the gravity component of the vibration data to obtain the target vibration data at any sampling time.

[0015] The average value of the target vibration data at each sampling moment within the time window is obtained to obtain the comprehensive vibration data at any monitoring moment.

[0016] Preferably, the step of obtaining the degree of influence of each type of environmental data based on the data fluctuation characteristics of multidimensional environmental data and comprehensive vibration data within a preset historical time period includes:

[0017] Multidimensional environmental data at each monitoring time within a preset historical period are combined into a multidimensional environmental data sequence. The multidimensional environmental data sequence is divided into at least two environmental data sequences according to the dimensions. For any environmental data sequence, the environmental data sequence is clustered to obtain at least one cluster.

[0018] For any two environmental data in any cluster, obtain the absolute value of the difference between the comprehensive vibration data at the monitoring time of the two environmental data, calculate the reciprocal of the absolute value of the difference and a preset constant, and obtain the degree of similarity between the comprehensive vibration data of the two environmental data.

[0019] The similarity of the comprehensive vibration data of each pair of environmental data in any cluster is obtained, and the average similarity of the comprehensive vibration data is obtained accordingly. The average value of the similarity of the average comprehensive vibration data for each cluster is obtained and normalized to obtain the influence degree of any environmental data.

[0020] Preferably, the step of obtaining normal comprehensive vibration data corresponding to each cluster based on the fluctuation characteristics of the vibration data corresponding to the monitoring time of the multidimensional environmental data in each cluster includes:

[0021] For any cluster, based on the fluctuation characteristics of the vibration data corresponding to the monitoring time of each multidimensional environmental data in the cluster, the reliability of the comprehensive vibration data at the monitoring time of each multidimensional environmental data in the cluster is obtained;

[0022] Obtain the cumulative confidence value of the comprehensive vibration data at the monitoring time of each multidimensional environmental data in any cluster. Calculate the ratio of the confidence value of the comprehensive vibration data at the monitoring time of each multidimensional environmental data to the cumulative confidence value of the comprehensive vibration data to obtain the weight of the comprehensive vibration data at the monitoring time of each multidimensional environmental data. Perform a weighted summation on the comprehensive vibration data at the monitoring time of each multidimensional environmental data in any cluster to obtain the normal comprehensive vibration data corresponding to any cluster.

[0023] Preferably, the step of obtaining the reliability of the comprehensive vibration data at the monitoring time corresponding to the vibration data at the monitoring time of each multidimensional environmental data in any cluster, based on the fluctuation characteristics of the vibration data at the monitoring time of each multidimensional environmental data in any cluster, includes:

[0024] For any monitoring time of any multidimensional environmental data in any cluster, within the time window corresponding to the monitoring time of any multidimensional environmental data, the standard deviation of the target vibration data at each sampling time is obtained. The reciprocal of the sum of the standard deviation and a preset constant is normalized to obtain the reliability of the comprehensive vibration data at the monitoring time of any multidimensional environmental data.

[0025] Preferably, the step of obtaining the degree of influence of environmental factors at the current monitoring time based on the difference between the multidimensional environmental data in each cluster and the multidimensional environmental data at the current monitoring time, the normal comprehensive vibration data corresponding to each cluster, and the degree of influence of each type of environmental data includes:

[0026] For any type of environmental data, based on the difference between the multidimensional environmental data in each cluster and the multidimensional environmental data at the current monitoring time, the degree of similarity between each cluster other than the aforementioned environmental data and other environments at the current monitoring time is obtained.

[0027] The mean environmental data belonging to any one type of environmental data in each cluster is obtained. The mean environmental data belonging to any one type of environmental data in each cluster is used as the horizontal axis and the normal comprehensive vibration data corresponding to each cluster is used as the vertical axis to construct a curve. The data points in the curve are fitted by the degree of similarity between each cluster other than the one type of environmental data and other environments at the current monitoring time to obtain a fitted curve.

[0028] Obtain the baseline environmental data for any one type of environmental data, use the fitting curve to obtain the fitting value of the baseline environmental data, and obtain the baseline normal comprehensive vibration data. Record the environmental data corresponding to any one type of environmental data at the current monitoring time as the current environmental data, use the fitting curve to obtain the fitting value of the current environmental data, and obtain the current normal comprehensive vibration data. Calculate the difference between the current normal comprehensive vibration data and the baseline normal comprehensive vibration data to obtain the difference value of the normal comprehensive vibration data for any one type of environmental data at the current monitoring time.

[0029] Obtain the difference value of normal comprehensive vibration data for each type of environmental data at the current monitoring time. Based on the difference value of normal comprehensive vibration data for each type of environmental data at the current monitoring time, and the degree of influence of each type of environmental data, obtain the degree of influence of environmental factors at the current monitoring time.

[0030] Preferably, the step of obtaining the degree of similarity between each cluster (excluding any one type of environmental data) and other environments at the current monitoring time based on the difference between the multidimensional environmental data in each cluster and the multidimensional environmental data at the current monitoring time includes:

[0031] For any cluster, the multidimensional environmental data in the cluster are formed into a multidimensional environmental data subsequence, and the multidimensional environmental data subsequence is divided into at least two environmental data subsequences according to the dimensions.

[0032] For any environmental data subsequence corresponding to any environmental data, environmental data subsequences other than the environmental data subsequence are denoted as other environmental data subsequences. For any other environmental data subsequence, environmental data of the same type as the other environmental data subsequence in the multidimensional environmental data at the current monitoring time are denoted as corresponding environmental data. The absolute value of the difference between the average value of the other environmental data subsequence and the corresponding environmental data is obtained to obtain the environmental difference value of the other environmental data subsequence.

[0033] Obtain the environmental difference value for each other environmental data subsequence, and obtain the corresponding cumulative environmental difference value. Normalize the reciprocal of the sum of the cumulative environmental difference value and the preset constant to obtain the degree of similarity between the cluster other than the environmental data and other environments at the current monitoring time.

[0034] Preferably, the step of obtaining the degree of influence of environmental factors at the current monitoring time based on the difference value of normal comprehensive vibration data for each type of environmental data at the current monitoring time, and the degree of influence of each type of environmental data, includes:

[0035] For any type of environmental data, the weighted influence degree of the environmental data is obtained by multiplying the difference value of the normal comprehensive vibration data of the environmental data at the current monitoring time with the influence degree of the environmental data.

[0036] The average weighted influence of each type of environmental data is obtained to determine the degree of environmental factor influence at the current monitoring time.

[0037] Preferably, the step of correcting each vibration data point corresponding to the current monitoring time using the degree of influence of the environmental factors to obtain a correction value for each vibration data point includes:

[0038] For any vibration data corresponding to the current monitoring time, the difference between the target vibration data at the sampling time of the vibration data and the degree of influence of the environmental factors is obtained, and the correction value of the vibration data is obtained.

[0039] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows:

[0040] In this invention, the degree of influence of each type of environmental data is obtained, reflecting the importance of each environmental factor on the vibration data. The greater the degree of influence, the greater the interference of the corresponding environmental factor on the vibration data. Normal comprehensive vibration data corresponding to each cluster is obtained, reflecting the vibration characteristics under different environmental conditions. This data is used to combine with multidimensional environmental data at the current monitoring time to obtain the degree of influence of environmental factors at the current monitoring time, analyze the influence relationship of multiple environmental factors on the tower vibration data under the combined action of multiple environmental factors, and correct the vibration data accordingly. Based on the corrected vibration data, the tower verticality is calculated, which can effectively reduce the interference of environmental factor changes on the monitoring results and improve the reliability of the online monitoring results of wind turbine tower verticality. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart of an online intelligent monitoring method for the verticality of a wind turbine tower provided in Embodiment 1 of the present invention. Detailed Implementation

[0043] Embodiments of this disclosure are described in detail below, with examples of these embodiments illustrated in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this disclosure, and should not be construed as limiting it.

[0044] It should be noted that the terms "first," "second," etc., used in this disclosure and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure.

[0045] To illustrate the technical solution of the present invention, specific embodiments are described below.

[0046] See Figure 1 This is a flowchart of an online intelligent monitoring method for the verticality of a wind turbine tower provided in Embodiment 1 of the present invention. Figure 1 As shown, the method may include:

[0047] Step S101: For any monitoring position in the wind turbine tower in any direction, based on the vibration data at each sampling time in the any direction, obtain the comprehensive vibration data at the current monitoring time and at each monitoring time within the preset historical period, and obtain the multidimensional environmental data at the current monitoring time and at each monitoring time within the preset historical period.

[0048] The wind turbine tower is the load-bearing structure supporting the nacelle and rotor of the wind turbine. Real-time monitoring of the tower's verticality is crucial for improving the safety and reliability of the wind turbine unit. Existing methods for monitoring wind turbine tower verticality mainly utilize monitoring equipment such as inclinometers and accelerometers to acquire tower vibration data in real time. The tower verticality is then calculated using this vibration data, and a fixed threshold method is used to remove some vibration data, or a simple linear model is used to compensate for environmental factors and reduce their influence. However, the combined effect of multiple environmental factors and the complexity of environmental changes make it difficult to accurately represent the environmental impact on tower vibration data using fixed thresholds or linear models, resulting in deviations in the final tower verticality readings.

[0049] Therefore, this embodiment of the invention compares the variation characteristics of wind turbine tower vibration data under different environmental factors to obtain the degree of influence of each environmental factor on the wind turbine tower vibration data. Then, it analyzes the degree of environmental factor influence on the tower vibration data under the combined action of multiple environmental factors, and corrects the tower vibration data accordingly. Based on the corrected data, the verticality of the wind turbine tower is calculated, which effectively reduces the interference of environmental factor changes on the monitoring results and improves the reliability of the online monitoring results of wind turbine tower verticality.

[0050] First, obtain vibration data of the wind turbine tower. Since it's necessary to calculate the overall verticality of the tower by considering the local tilt angles at multiple locations on the tower, vibration data is collected at various points on the tower. high, high, Monitoring positions are set at the height and top of the tower. Bidirectional accelerometers are arranged at each monitoring position to collect vibration data (i.e., acceleration data) in two orthogonal directions of the tower cross section. In this embodiment, the setting of monitoring positions is not limited and can be set according to the specific implementation scenario.

[0051] Since the verticality of the wind turbine tower is mainly determined by the vibration characteristics analysis of vibration data, the vibration data acquisition frequency is set to 100Hz in this embodiment to accurately reflect the vibration characteristics. This setting is not limited and can be adjusted according to the specific implementation scenario. Because the acceleration data collected by the accelerometer is affected by gravitational acceleration, a tilt sensor (triaxial tiltmeter) is installed at the top of the wind turbine tower. The z-axis is the vertical direction, and the x and y axes are consistent with the monitoring direction of the accelerometer. Since the tilt angle of the wind turbine tower changes relatively slowly, the monitoring frequency of the tilt sensor is set to 1Hz in this embodiment. This setting is not limited and can be adjusted according to the specific implementation scenario.

[0052] Then, multi-dimensional environmental data of the wind turbine tower is acquired. The main environmental factors affecting the vibration of the wind turbine tower are temperature, humidity and wind speed. Therefore, a thermometer, humidity sensor, anemometer and wind direction meter are arranged at each acceleration sensor position to monitor the temperature data, humidity data, wind speed data and wind direction data at each position. The environment where the wind turbine tower is located changes relatively slowly, so the monitoring frequency of environmental data in this embodiment is set to 1Hz. This is not limited here and can be set according to the specific implementation scenario.

[0053] To facilitate the analysis of the relationship between different data, the collected vibration data, tilt data and environmental data need to be normalized. The data mentioned later in this embodiment are all normalized data.

[0054] Since the analysis process is consistent across all monitoring locations and directions within the wind turbine tower, this embodiment uses any monitoring location and any direction within the wind turbine tower as an example. Based on the vibration data at each sampling moment in that direction, comprehensive vibration data is obtained at the current monitoring moment and at each monitoring moment within a preset historical period. Multidimensional environmental data (temperature, humidity, wind speed, and wind direction) are also obtained at the current monitoring moment and at each monitoring moment within the preset historical period to analyze the impact of the environment on the vibration data. Because short-term environmental conditions are closer to the current monitoring moment and have higher reference value, and to avoid excessive data burden on the system analysis, this embodiment sets the preset historical period to 3 days prior to the current monitoring moment. This is not limited and can be set according to the specific implementation scenario.

[0055] The method for obtaining comprehensive vibration data at the current monitoring time and at each monitoring time within a preset historical period based on the vibration data at each sampling time in any direction is as follows:

[0056] In order to align the collected vibration data with the monitoring time of other monitoring data, a time window of a preset length is constructed for the current monitoring time and any monitoring time within the preset historical period, with the any monitoring time as the cutoff time. The preset length is the interval between the any monitoring time and its previous monitoring time, that is, one time window is obtained per second.

[0057] Since the acceleration data collected by the accelerometer is affected by gravitational acceleration, it is necessary to remove the influence of gravitational acceleration. The tilt angle monitoring data collected by the tilt angle sensor at any given monitoring time is obtained. For any sampling time within the time window, the gravity component of the vibration data at that sampling time (i.e., the component of gravitational acceleration in the monitoring direction) is obtained using the tilt angle monitoring data. The difference between the vibration data at that sampling time and the gravity component of the vibration data is calculated to obtain the target vibration data at that sampling time (i.e., the vibration data after removing gravitational interference). Obtaining the gravity component of vibration data using tilt angle monitoring data is existing technology and will not be elaborated here.

[0058] Since the calculation of tower verticality focuses more on the trend of vibration data and needs to reduce the vibration impact of vibration data, the average value of the target vibration data at each sampling moment within the time window is obtained to obtain the comprehensive vibration data at any monitoring moment, that is, the overall vibration data of the time window corresponding to any monitoring moment.

[0059] Similarly, comprehensive vibration data is obtained at each monitoring time.

[0060] Thus, vibration data at each sampling time in any direction is obtained, along with comprehensive vibration data and multidimensional environmental data at the current monitoring time and at each monitoring time within its preset historical period.

[0061] Step S102: Based on the data fluctuation characteristics of multidimensional environmental data and comprehensive vibration data within a preset historical period, obtain the influence degree of each type of environmental data. Using the influence degree of each type of environmental data, cluster the multidimensional environmental data within the preset historical period to obtain at least one cluster. Based on the fluctuation characteristics of the vibration data corresponding to the monitoring time of the multidimensional environmental data in each cluster, obtain the normal comprehensive vibration data corresponding to each cluster.

[0062] Since the impact of environmental factors on tower vibration data is not a simple sum of the effects of various environmental factors, in order to obtain a more accurate relationship between the impact of environmental factors on tower vibration data, it is necessary to integrate multiple environmental factors at each monitoring time, and analyze the relationship between the impact of environmental factors on tower vibration data by comparing the changes in characteristic data under each environmental condition.

[0063] Because different environmental factors may have varying degrees of influence on tower vibration data, the importance of different environmental factors is not consistent when analyzing their impact on tower vibration data. If, at different monitoring times, a certain environmental factor is similar and the difference in the overall vibration data is small, it indicates that other environmental factors besides that factor have a smaller impact on the overall vibration data; that is, the environmental factor has a larger impact on the overall vibration data, and its importance is greater. Therefore, the degree of influence of each environmental factor can be obtained based on the data fluctuation characteristics of multidimensional environmental data and overall vibration data within a preset historical period, reflecting the importance of each environmental factor on the vibration data.

[0064] The method for obtaining the degree of influence of each type of environmental data based on the data fluctuation characteristics of multidimensional environmental data and comprehensive vibration data within a preset historical time period is as follows:

[0065] Multidimensional environmental data at each monitoring time within a preset historical period are combined into a multidimensional environmental data sequence. The multidimensional environmental data sequence is divided into at least two environmental data sequences according to the dimensions. For any environmental data sequence, the k-means clustering method is used to cluster the environmental data sequence to obtain at least one cluster. k-means clustering is an existing technology and will not be elaborated here.

[0066] For any two environmental data in any cluster, obtain the absolute value of the difference between the comprehensive vibration data at the monitoring time of the two environmental data, calculate the reciprocal of the absolute value of the difference and a preset constant, and obtain the degree of similarity between the comprehensive vibration data of the two environmental data.

[0067] The similarity of the comprehensive vibration data of each pair of environmental data in any cluster is obtained, and the average similarity of the comprehensive vibration data is obtained accordingly. The average value of the similarity of the average comprehensive vibration data for each cluster is obtained and normalized to obtain the influence degree of any environmental data.

[0068] In one embodiment, taking the u-th type of environmental data as an example, the formula for calculating the degree of influence of the u-th type of environmental data is:

[0069]

[0070] in, The degree of impact of the u-th type of environmental data; After clustering the data sequence corresponding to the u-th type of environmental data, the absolute value of the difference between the comprehensive vibration data of the i-th environmental data and the j-th environmental data at the monitoring time in the w-th cluster; The number of environmental data in the w-th cluster after clustering the data sequence corresponding to the u-th type of environmental data; The number of clusters obtained after clustering the data sequence corresponding to the u-th type of environmental data; This is the normalization function; As a preset constant, this embodiment sets This is used to ensure that the fraction is meaningful. There are no restrictions here, and it can be set according to the specific implementation scenario.

[0071] It should be noted that after clustering the data sequences corresponding to the u-th type of environmental data, the environmental data in each cluster are environmental data under similar environmental factor conditions. The smaller the value, the more similar the integrated vibration data at the monitoring time of the i-th and j-th environmental data points in the w-th cluster. In other words, under similar environmental conditions, the difference in integrated vibration data at different monitoring times is smaller, and the importance of the u-th environmental data point is greater. The larger it is.

[0072] Following the method for obtaining the influence degree of the u-th type of environmental data described above, the influence degree of each type of environmental data is obtained. Since tower vibration data may generate noise data due to interference factors such as wind turbine start-up and shutdown, and abnormal data transmission, this noise data cannot reflect the influence relationship of environmental factors on tower vibration data. Therefore, in this embodiment, the influence degree of each type of environmental data is used as the weight of the corresponding environmental data in each multidimensional environmental data within a preset historical period. The k-means clustering method is used to perform weighted clustering on the multidimensional environmental data within the preset historical period to obtain at least one cluster. The environmental conditions at the monitoring time of the multidimensional environmental data in each cluster are similar. k-means clustering is an existing technology and will not be elaborated here. Then, based on the fluctuation characteristics of the vibration data corresponding to the monitoring time of the multidimensional environmental data in each cluster, the normal comprehensive vibration data corresponding to each cluster is obtained, reflecting the vibration characteristics under different environmental factors.

[0073] The method for obtaining the normal comprehensive vibration data for each cluster, based on the fluctuation characteristics of the vibration data corresponding to the monitoring time of the multidimensional environmental data in each cluster, is as follows:

[0074] (1) For any cluster, based on the fluctuation characteristics of the vibration data corresponding to the monitoring time of each multidimensional environmental data in the cluster, obtain the reliability of the comprehensive vibration data at the monitoring time of each multidimensional environmental data in the cluster.

[0075] Specifically, since the comprehensive vibration data at each monitoring moment is the average target vibration data within the corresponding time window, it may be affected by noise interference, leading to significant differences in comprehensive vibration data under the same environmental conditions. To reduce the impact of noise-affected data on the analysis results, for any monitoring moment of any multidimensional environmental data in any cluster, the standard deviation of the target vibration data at each sampling moment is obtained within the time window corresponding to the monitoring moment of any multidimensional environmental data. The reciprocal of the sum of the standard deviation and a preset constant is normalized to obtain the reliability of the comprehensive vibration data at the monitoring moment of any multidimensional environmental data.

[0076] In one embodiment, taking the monitoring time of the kth multidimensional environmental data point as an example, the formula for calculating the reliability of the comprehensive vibration data at the monitoring time of the kth multidimensional environmental data point is as follows:

[0077]

[0078] in, The reliability of the comprehensive vibration data at the monitoring time of the k-th multidimensional environmental data point; For the time window corresponding to the monitoring time of the k-th multidimensional environmental data, the standard deviation of the target vibration data at each sampling time. This is the normalization function; As a preset constant, this embodiment sets This is used to ensure that the fraction is meaningful. There are no restrictions here, and it can be set according to the specific implementation scenario.

[0079] It should be noted that, The smaller the value, the more stable the vibration data is within the time window corresponding to the monitoring time of the k-th multidimensional environmental data point, meaning there is less likelihood of interference data caused by sudden noise changes. The larger it is.

[0080] Similarly, the reliability of the comprehensive vibration data at the monitoring time of each multidimensional environmental data in any of the aforementioned clusters is obtained.

[0081] (2) Based on the reliability of the comprehensive vibration data at the monitoring time of each multidimensional environmental data in any cluster, obtain the normal comprehensive vibration data corresponding to any cluster.

[0082] Specifically, the cumulative confidence level of the comprehensive vibration data at the monitoring time of each multidimensional environmental data in any cluster is obtained. The ratio of the confidence level of the comprehensive vibration data at the monitoring time of each multidimensional environmental data to the cumulative confidence level of the comprehensive vibration data is calculated to obtain the weight of the comprehensive vibration data at the monitoring time of each multidimensional environmental data. The comprehensive vibration data at the monitoring time of each multidimensional environmental data in any cluster is then weighted and summed to obtain the normal comprehensive vibration data corresponding to any cluster.

[0083] In one embodiment, taking the v-th cluster as an example, the formula for calculating the normal composite vibration data corresponding to the v-th cluster is:

[0084]

[0085] in, This refers to the normal comprehensive vibration data corresponding to the v-th cluster; The reliability of the comprehensive vibration data at the monitoring time of the s-th multidimensional environmental data point in the v-th cluster; This refers to the comprehensive vibration data at the monitoring time of the s-th multidimensional environmental data point in the v-th cluster; denoted as the number of multidimensional environmental data in the v-th cluster.

[0086] It should be noted that, The larger, the more The more likely the data is to be normal, integrated vibration data free from noise interference, the better the calculated values ​​will be. The more credible it is, the more believable it becomes.

[0087] Similarly, the normal comprehensive vibration data corresponding to each cluster is obtained.

[0088] Step S103: Based on the difference between the multidimensional environmental data in each cluster and the multidimensional environmental data at the current monitoring time, the normal comprehensive vibration data corresponding to each cluster, and the degree of influence of each type of environmental data, obtain the degree of influence of environmental factors at the current monitoring time, and use the degree of influence of environmental factors to correct each vibration data corresponding to the current monitoring time to obtain the correction value of each vibration data.

[0089] After obtaining the normal comprehensive vibration data corresponding to each cluster, based on the difference between the multidimensional environmental data in each cluster and the multidimensional environmental data at the current monitoring time, the normal comprehensive vibration data corresponding to each cluster, and the degree of influence of each type of environmental data, the degree of influence of environmental factors at the current monitoring time is obtained. The influence relationship of multiple environmental factors on the tower vibration data under the combined action is analyzed, and the tower vibration data is corrected accordingly.

[0090] The method for obtaining the degree of influence of environmental factors at the current monitoring time is as follows, based on the difference between the multidimensional environmental data in each cluster and the multidimensional environmental data at the current monitoring time, the normal comprehensive vibration data corresponding to each cluster, and the degree of influence of each type of environmental data:

[0091] (1) For any type of environmental data, based on the difference between the multidimensional environmental data in each cluster and the multidimensional environmental data at the current monitoring time, obtain the degree of similarity between each cluster other than the aforementioned environmental data and other environments at the current monitoring time.

[0092] Specifically, for any cluster, the multidimensional environmental data in the cluster are grouped into a multidimensional environmental data subsequence, and the multidimensional environmental data subsequence is divided into at least two environmental data subsequences according to the dimensions.

[0093] For any environmental data subsequence corresponding to any environmental data, environmental data subsequences other than the environmental data subsequence are denoted as other environmental data subsequences. For any other environmental data subsequence, environmental data of the same type as the other environmental data subsequence in the multidimensional environmental data at the current monitoring time are denoted as corresponding environmental data. The absolute value of the difference between the average value of the other environmental data subsequence and the corresponding environmental data is obtained to obtain the environmental difference value of the other environmental data subsequence.

[0094] Obtain the environmental difference value for each other environmental data subsequence, and obtain the corresponding cumulative environmental difference value. Normalize the reciprocal of the sum of the cumulative environmental difference value and the preset constant to obtain the degree of similarity between the cluster other than the environmental data and other environments at the current monitoring time.

[0095] In one embodiment, taking the v-th cluster and the u-th environmental data as an example, the formula for calculating the similarity between the v-th cluster (excluding the u-th environmental data) and other environments at the current monitoring time is as follows:

[0096]

[0097] in, This represents the degree of similarity between the v-th cluster (excluding the u-th environmental data) and other environments at the current monitoring time. It represents the average value of the r-th other environmental data subsequence in the v-th cluster; The environmental data in the multidimensional environmental data at the current monitoring time is of the same type as the r-th other environmental data subsequence; N is the number of environmental data subsequences, that is, the total number of environmental data types; This is the normalization function; As a preset constant, this embodiment sets This is used to ensure that the fraction is meaningful. There are no restrictions here, and it can be set according to the specific implementation scenario. It is the absolute value symbol.

[0098] It should be noted that, The smaller the value, the more similar the environmental conditions at the current monitoring time are to the environmental conditions corresponding to the v-th cluster, excluding the u-th type of environmental data. The larger it is.

[0099] Similarly, the proximity of each cluster to other environments at the current monitoring time is obtained, in addition to any of the aforementioned environmental data.

[0100] (2) Obtain the difference value of normal comprehensive vibration data for each type of environmental data at the current monitoring time.

[0101] Specifically, the mean environmental data belonging to any one type of environmental data in each cluster is obtained. A curve is constructed with the mean environmental data belonging to any one type of environmental data in each cluster as the x-axis and the normal comprehensive vibration data corresponding to each cluster as the y-axis. The degree of similarity between each cluster other than any one type of environmental data and other environments at the current monitoring time is used as the weight. The data points in the curve are fitted using the weighted least squares method to obtain the fitted curve. The weighted least squares method is an existing technology and will not be described in detail here.

[0102] The baseline environmental data for obtaining any of the aforementioned environmental data is as follows: Since the temperature and humidity may vary significantly in different regions under normal conditions, this embodiment uses the median temperature data of the region where the wind turbine tower is located over three years as the baseline temperature data, and the median humidity data of the region where the wind turbine tower is located over three years as the baseline humidity data. Since wind speed usually has a significant impact on the wind turbine tower, this embodiment sets the baseline wind speed data to the wind speed data under no-wind conditions, i.e., 0 m / s. The average wind direction data of the region where the wind turbine tower is located over three years is used as the baseline wind direction data. This is not limited here and can be set according to the specific implementation scenario.

[0103] Substitute the baseline environmental data of any one of the environmental data types into the fitting curve to obtain the fitting value of the baseline environmental data, thus obtaining the baseline normal comprehensive vibration data. Record the environmental data corresponding to any one of the environmental data types at the current monitoring time as the current environmental data. Substitute the current environmental data into the fitting curve to obtain the fitting value of the current environmental data, thus obtaining the current normal comprehensive vibration data. Calculate the difference between the current normal comprehensive vibration data and the baseline normal comprehensive vibration data to obtain the difference value of the normal comprehensive vibration data of any one of the environmental data types at the current monitoring time.

[0104] Similarly, obtain the difference value of normal comprehensive vibration data for each type of environmental data at the current monitoring time.

[0105] (3) Based on the difference in normal comprehensive vibration data of each type of environmental data at the current monitoring time, and the degree of influence of each type of environmental data, obtain the degree of influence of environmental factors at the current monitoring time.

[0106] Specifically, for any type of environmental data, the product between the difference value of the normal comprehensive vibration data of the environmental data at the current monitoring time and the influence degree of the environmental data is obtained to obtain the weighted influence degree of the environmental data.

[0107] The average weighted influence of each type of environmental data is obtained to determine the degree of environmental factor influence at the current monitoring time.

[0108] In one embodiment, the formula for calculating the degree of influence of environmental factors at the current monitoring time is:

[0109]

[0110] in, The degree of influence of environmental factors at the current monitoring time; Let be the difference value of the normal comprehensive vibration data of the u-th environmental data at the current monitoring time; denoted by u, representing the degree of influence of the u-th type of environmental data; N represents the total number of types of environmental data.

[0111] It should be noted that, and The larger the value, the greater the impact of the u-th type of environmental data on the overall vibration data at the current monitoring time. The larger it is.

[0112] Furthermore, the degree of influence of environmental factors at the current monitoring time is used to correct each vibration data point corresponding to the current monitoring time, resulting in a correction value for each vibration data point. Specifically:

[0113] For any vibration data corresponding to the current monitoring time, the difference between the target vibration data at the sampling time of the vibration data and the degree of influence of the environmental factors is obtained, and the correction value of the vibration data is obtained.

[0114] In one embodiment, taking the t-th vibration data point corresponding to the current monitoring time as an example, the formula for calculating the correction value of the t-th vibration data point is:

[0115]

[0116] in, This is the correction value for the t-th vibration data point; This is the t-th vibration data point corresponding to the current monitoring time; This represents the degree of influence of environmental factors at the current monitoring time.

[0117] Following the method described above for obtaining the correction value of the t-th vibration data point, obtain the correction value for each vibration data point corresponding to the current monitoring time.

[0118] Thus, the correction value for each vibration data point corresponding to the current monitoring time is obtained.

[0119] Step S104: Obtain the verticality of the wind turbine tower by using the correction value of each vibration data corresponding to each direction at each monitoring position in the wind turbine tower at the current monitoring time.

[0120] Following the method described above for obtaining the correction value of each vibration data point at the current monitoring time, the correction value of each vibration data point in each direction at each monitoring location within the wind turbine tower is obtained at the current monitoring time. Then, using the correction values ​​of each vibration data point in each direction at each monitoring location within the wind turbine tower at the current monitoring time, the verticality of the wind turbine tower is calculated. This effectively reduces the interference of environmental factors on the monitoring results and improves the reliability of the online monitoring results for the verticality of the wind turbine tower.

[0121] Calculating wind turbine tower verticality using tower vibration data is existing technology, which will be briefly described here:

[0122] The correction value of each vibration data point at each monitoring location in each direction at the current monitoring time is processed by second-order integration to obtain the displacement of each monitoring location in each direction, thereby generating the three-dimensional coordinates of each monitoring location (i.e., displacement in the x-direction, displacement in the y-direction, and height of the monitoring location). The three-dimensional coordinates of all monitoring locations are fitted using the least squares method to obtain the fitted curve as the tower axis at the current monitoring time. Based on the fitted curve, the fitted displacement in the x-direction and y-direction of the monitoring location at the top of the tower is obtained. The x-direction and y-direction coordinates of the monitoring location at the top of the tower in a completely vertical and unoffset state are marked as the reference coordinates (0, 0). The difference between the fitted displacement in the x-direction and y-direction of the monitoring location at the top of the tower and the reference coordinates is calculated to obtain the offset in the x-direction ∆x and the offset in the y-direction ∆y. Then, the verticality of the wind turbine tower at the current monitoring time is calculated. .

[0123] In summary, in this embodiment of the invention, the degree of influence of each type of environmental data is obtained, reflecting the importance of each environmental factor on the vibration data. The greater the degree of influence, the greater the interference of the corresponding environmental factor on the vibration data. Normal comprehensive vibration data corresponding to each cluster is obtained, reflecting the vibration characteristics under different environmental conditions. This data is used to combine with the multidimensional environmental data at the current monitoring time to obtain the degree of influence of environmental factors at the current monitoring time, analyze the influence relationship of multiple environmental factors on the tower vibration data under the combined action of multiple environmental factors, and correct the vibration data accordingly. Based on the corrected vibration data, the tower verticality is calculated, which can effectively reduce the interference of environmental factor changes on the monitoring results and improve the reliability of the online monitoring results of wind turbine tower verticality.

[0124] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. An online intelligent monitoring method for the verticality of wind turbine towers, characterized in that, include: For any monitoring position in the wind turbine tower in any direction, based on the vibration data at each sampling time in any direction, obtain the comprehensive vibration data at the current monitoring time and at each monitoring time within the preset historical period, and obtain the multi-dimensional environmental data at the current monitoring time and at each monitoring time within the preset historical period. Based on the data fluctuation characteristics of multidimensional environmental data and comprehensive vibration data within a preset historical period, the influence degree of each type of environmental data is obtained. Using the influence degree of each type of environmental data, the multidimensional environmental data within the preset historical period is clustered to obtain at least one cluster. Based on the fluctuation characteristics of the vibration data corresponding to the monitoring time of the multidimensional environmental data in each cluster, the normal comprehensive vibration data corresponding to each cluster is obtained. Based on the difference between the multidimensional environmental data in each cluster and the multidimensional environmental data at the current monitoring time, the normal comprehensive vibration data corresponding to each cluster, and the degree of influence of each type of environmental data, the degree of influence of environmental factors at the current monitoring time is obtained. The degree of influence of environmental factors is used to correct each vibration data corresponding to the current monitoring time to obtain the correction value of each vibration data. The verticality of the wind turbine tower is obtained by using the correction value of each vibration data corresponding to each direction at each monitoring position in the wind turbine tower at the current monitoring time; The step of obtaining comprehensive vibration data for the current monitoring time and for each monitoring time within a preset historical period based on vibration data at each sampling time in any direction includes: For the current monitoring time and any monitoring time within a preset historical period, a time window of a preset length is constructed with the any monitoring time as the cutoff time. The preset length is the interval between the any monitoring time and its previous monitoring time. Acquire tilt monitoring data at any monitoring time. For any sampling time within the time window, use the tilt monitoring data to obtain the gravity component of the vibration data at any sampling time. Calculate the difference between the vibration data at any sampling time and the gravity component of the vibration data to obtain the target vibration data at any sampling time. The average value of the target vibration data at each sampling moment within the time window is obtained to obtain the comprehensive vibration data at any monitoring moment. The process of obtaining the degree of influence of each type of environmental data based on the data fluctuation characteristics of multidimensional environmental data and comprehensive vibration data within a preset historical time period includes: Multidimensional environmental data at each monitoring time within a preset historical period are combined into a multidimensional environmental data sequence. The multidimensional environmental data sequence is divided into at least two environmental data sequences according to the dimensions. For any environmental data sequence, the environmental data sequence is clustered to obtain at least one cluster. For any two environmental data in any cluster, obtain the absolute value of the difference between the comprehensive vibration data at the monitoring time of the two environmental data, calculate the reciprocal of the absolute value of the difference and a preset constant, and obtain the degree of similarity between the comprehensive vibration data of the two environmental data. The similarity of the comprehensive vibration data of each pair of environmental data in any cluster is obtained, and the average similarity of the comprehensive vibration data is obtained accordingly. The average value of the average similarity of the comprehensive vibration data for each cluster is obtained and normalized to obtain the influence degree of any environmental data.

2. The online intelligent monitoring method for the verticality of a wind turbine tower according to claim 1, characterized in that, The process of obtaining normal comprehensive vibration data for each cluster based on the fluctuation characteristics of the vibration data corresponding to the monitoring time of the multidimensional environmental data in each cluster includes: For any cluster, based on the fluctuation characteristics of the vibration data corresponding to the monitoring time of each multidimensional environmental data in the cluster, the reliability of the comprehensive vibration data at the monitoring time of each multidimensional environmental data in the cluster is obtained; Obtain the cumulative confidence value of the comprehensive vibration data at the monitoring time of each multidimensional environmental data in any cluster. Calculate the ratio of the confidence value of the comprehensive vibration data at the monitoring time of each multidimensional environmental data to the cumulative confidence value of the comprehensive vibration data to obtain the weight of the comprehensive vibration data at the monitoring time of each multidimensional environmental data. Perform a weighted summation on the comprehensive vibration data at the monitoring time of each multidimensional environmental data in any cluster to obtain the normal comprehensive vibration data corresponding to any cluster.

3. The online intelligent monitoring method for the verticality of a wind turbine tower according to claim 2, characterized in that, The step of obtaining the reliability of the comprehensive vibration data at the monitoring time corresponding to the vibration data at the monitoring time of each multidimensional environmental data in any cluster, based on the fluctuation characteristics of the vibration data at the monitoring time of each multidimensional environmental data in any cluster, includes: For any monitoring time of any multidimensional environmental data in any cluster, within the time window corresponding to the monitoring time of any multidimensional environmental data, the standard deviation of the target vibration data at each sampling time is obtained. The reciprocal of the sum of the standard deviation and a preset constant is normalized to obtain the reliability of the comprehensive vibration data at the monitoring time of any multidimensional environmental data.

4. The online intelligent monitoring method for the verticality of a wind turbine tower according to claim 1, characterized in that, The method of obtaining the degree of influence of environmental factors at the current monitoring time based on the difference between the multidimensional environmental data in each cluster and the multidimensional environmental data at the current monitoring time, the normal comprehensive vibration data corresponding to each cluster, and the degree of influence of each type of environmental data includes: For any type of environmental data, based on the difference between the multidimensional environmental data in each cluster and the multidimensional environmental data at the current monitoring time, the degree of similarity between each cluster other than the aforementioned environmental data and other environments at the current monitoring time is obtained. The mean environmental data belonging to any one type of environmental data in each cluster is obtained. The mean environmental data belonging to any one type of environmental data in each cluster is used as the horizontal axis and the normal comprehensive vibration data corresponding to each cluster is used as the vertical axis to construct a curve. The data points in the curve are fitted by the degree of similarity between each cluster other than the one type of environmental data and other environments at the current monitoring time to obtain a fitted curve. Obtain the baseline environmental data for any one type of environmental data, use the fitting curve to obtain the fitting value of the baseline environmental data, and obtain the baseline normal comprehensive vibration data. Record the environmental data corresponding to any one type of environmental data at the current monitoring time as the current environmental data, use the fitting curve to obtain the fitting value of the current environmental data, and obtain the current normal comprehensive vibration data. Calculate the difference between the current normal comprehensive vibration data and the baseline normal comprehensive vibration data to obtain the difference value of the normal comprehensive vibration data for any one type of environmental data at the current monitoring time. Obtain the difference value of normal comprehensive vibration data for each type of environmental data at the current monitoring time. Based on the difference value of normal comprehensive vibration data for each type of environmental data at the current monitoring time, and the degree of influence of each type of environmental data, obtain the degree of influence of environmental factors at the current monitoring time.

5. The online intelligent monitoring method for the verticality of a wind turbine tower according to claim 4, characterized in that, The step of obtaining the degree of similarity between each cluster and other environments at the current monitoring time, excluding any one type of environmental data, based on the difference between the multidimensional environmental data in each cluster and the multidimensional environmental data at the current monitoring time includes: For any cluster, the multidimensional environmental data in the cluster are formed into a multidimensional environmental data subsequence, and the multidimensional environmental data subsequence is divided into at least two environmental data subsequences according to the dimensions. For any environmental data subsequence corresponding to any environmental data, environmental data subsequences other than the environmental data subsequence are denoted as other environmental data subsequences. For any other environmental data subsequence, environmental data of the same type as the other environmental data subsequence in the multidimensional environmental data at the current monitoring time are denoted as corresponding environmental data. The absolute value of the difference between the average value of the other environmental data subsequence and the corresponding environmental data is obtained to obtain the environmental difference value of the other environmental data subsequence. Obtain the environmental difference value for each other environmental data subsequence, and obtain the corresponding cumulative environmental difference value. Normalize the reciprocal of the sum of the cumulative environmental difference value and the preset constant to obtain the degree of similarity between the cluster other than the environmental data and other environments at the current monitoring time.

6. The online intelligent monitoring method for the verticality of a wind turbine tower according to claim 4, characterized in that, The process of obtaining the degree of influence of environmental factors at the current monitoring time based on the difference in normal comprehensive vibration data for each type of environmental data at the current monitoring time, and the degree of influence of each type of environmental data, includes: For any type of environmental data, the weighted influence degree of the environmental data is obtained by multiplying the difference value of the normal comprehensive vibration data of the environmental data at the current monitoring time with the influence degree of the environmental data. The average weighted influence of each type of environmental data is obtained to determine the degree of environmental factor influence at the current monitoring time.

7. The online intelligent monitoring method for the verticality of a wind turbine tower according to claim 1, characterized in that, The step of correcting each vibration data point at the current monitoring time using the degree of influence of the environmental factors to obtain a correction value for each vibration data point includes: For any vibration data corresponding to the current monitoring time, the difference between the target vibration data at the sampling time of the vibration data and the degree of influence of the environmental factors is obtained, and the correction value of the vibration data is obtained.

Images