A loosening early warning method for a fan tower fastening bolt

Abstract

The present application relates to the technical field of data processing, and especially relates to a loosening early warning method for a fastening bolt of a fan tower, which obtains multi-dimensional monitoring data of any fastening bolt in the fan tower at each time, obtains a multi-dimensional monitoring data sequence containing the current time, obtains a bolt suspected loosening degree according to fluctuation characteristics of the multi-dimensional monitoring data sequence, obtains a bolt loosening confidence according to data differences between the multi-dimensional monitoring data at the current time and the multi-dimensional monitoring data at the last time and variation trend characteristics of each dimension monitoring data in the multi-dimensional monitoring data sequence, obtains an aging degree of any fastening bolt, obtains a bolt loosening degree according to the bolt suspected loosening degree, the bolt loosening confidence and the aging degree of any fastening bolt, adjusts a preset early warning threshold according to the bolt loosening degree, obtains an adaptive early warning threshold of any fastening bolt at the current time, carries out loosening early warning on any fastening bolt, and improves early warning accuracy.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method for early warning of loosening of wind turbine tower fastening bolts. Background Technology

[0002] With the rapid development of the wind power generation industry, wind turbine generators (wind turbines) have been widely used in renewable energy projects around the world. As a crucial component of the wind turbine, the tower bears the weight and wind load of the turbine, ensuring its stability during long-term operation. Among these components, the tower's fastening bolts play a critical role, securing various parts and ensuring the overall stability and safety of the tower. However, during long-term operation, wind turbine towers are subjected to complex wind loads, temperature changes, vibrations, and environmental factors (such as humidity and salt spray), which can cause these fastening bolts to loosen. This loosening not only affects the structural stability of the tower but may even lead to serious safety accidents. Therefore, using relevant data on the wind turbine tower's fastening bolts for early warning of loosening is essential.

[0003] Traditional methods for early warning of loosening wind turbine tower bolts typically employ vibration signal analysis techniques with fixed thresholds. For example, they monitor the root mean square (RMS) value of the vibration signal and compare it to a preset threshold to determine if the tower bolts are loose, then issue an early warning. However, wind turbine operating conditions are complex (e.g., wind speed fluctuations, temperature changes, equipment aging), and environmental changes can cause normal vibration amplitudes to exceed the fixed threshold, resulting in a high false alarm rate. Furthermore, when bolts loosen slowly, the vibration amplitude change may not reach the fixed threshold, leading to a high risk of missed alarms. Additionally, traditional methods cannot adapt to equipment aging or load changes throughout the wind turbine's lifespan, making early warnings of loosening tower bolts inaccurate under fixed thresholds.

[0004] Therefore, how to adaptively adjust the warning threshold of the wind turbine tower fastening bolts according to the actual operating conditions of the wind turbine and improve the accuracy of the warning has become an urgent problem to be solved. Summary of the Invention

[0005] In view of this, embodiments of the present invention provide a method for early warning of loosening of wind turbine tower fastening bolts, in order to solve the problem of how to adaptively adjust the early warning threshold of wind turbine tower fastening bolts according to the actual operating conditions of the wind turbine and improve the accuracy of the early warning.

[0006] This invention provides a method for early warning of loosening of wind turbine tower fastening bolts, the method comprising the following steps:

[0007] For any fastening bolt in the wind turbine tower, according to a preset monitoring frequency, multidimensional monitoring data of the fastening bolt at each moment is acquired to obtain a multidimensional monitoring data sequence within a preset time period including the current moment. The multidimensional monitoring data includes vibration data, wind speed data, temperature data, and torque data.

[0008] Based on the fluctuation characteristics of the vibration data in the multidimensional monitoring data sequence, the suspected looseness of any fastening bolt at the current moment is obtained. Based on the data difference between the multidimensional monitoring data at the current moment and the multidimensional monitoring data at the previous moment, as well as the changing trend characteristics of the monitoring data in each dimension of the multidimensional monitoring data sequence, the bolt looseness confidence of any fastening bolt at the current moment is obtained.

[0009] Obtain the aging degree of any of the fastening bolts, and based on the suspected loosening degree of the bolt, the loosening confidence of the bolt, and the aging degree of any of the fastening bolts, obtain the bolt loosening degree of any of the fastening bolts at the current moment;

[0010] Based on the degree of looseness of any fastening bolt at the current moment, the preset warning threshold is adjusted to obtain the adaptive warning threshold of any fastening bolt at the current moment. Based on the difference between the vibration data at the current moment and the adaptive warning threshold, a loosening warning is issued for any fastening bolt.

[0011] Preferably, the step of obtaining the suspected looseness of any fastening bolt at the current moment based on the fluctuation characteristics of the vibration data in the multidimensional monitoring data sequence includes:

[0012] The absolute value of the difference between the vibration data at the current moment and the vibration data at the previous moment is obtained to obtain the instantaneous fluctuation value. The vibration data in the multidimensional monitoring data sequence is fitted to obtain a first-order vibration fitting function. The slope of the first-order vibration fitting function is obtained. The slope is normalized to obtain the overall fluctuation value. The product of the instantaneous fluctuation value and the overall fluctuation value is obtained to obtain the suspected looseness of any fastening bolt at the current moment.

[0013] Preferably, the step of obtaining the bolt loosening confidence level of any fastening bolt at the current moment based on the data difference between the multidimensional monitoring data at the current moment and the multidimensional monitoring data at the previous moment, and the changing trend characteristics of the monitoring data in each dimension of the multidimensional monitoring data sequence, includes:

[0014] Based on the data difference between the current multidimensional monitoring data and the previous multidimensional monitoring data, the first bolt loosening confidence level of any fastening bolt at the current moment is obtained;

[0015] Based on the changing trend characteristics of the monitoring data in each dimension of the multidimensional monitoring data sequence, the second bolt loosening confidence level of any fastening bolt at the current moment is obtained;

[0016] The bolt loosening confidence level of any fastening bolt at the current moment is obtained by averaging the first bolt loosening confidence level and the second bolt loosening confidence level.

[0017] Preferably, obtaining the first bolt loosening confidence level of any fastening bolt at the current moment based on the data difference between the multidimensional monitoring data at the current moment and the multidimensional monitoring data at the previous moment includes:

[0018] Obtain the absolute value of the difference between the wind speed data at the current moment and the wind speed data at the previous moment to obtain the wind speed difference value; obtain the absolute value of the difference between the temperature data at the current moment and the temperature data at the previous moment to obtain the temperature difference value; obtain the product of the wind speed difference value and the temperature difference value to obtain the environmental change value; obtain the reciprocal of the sum of the environmental change value and the preset constant to obtain the environmental stability level.

[0019] The difference between the torque data of the previous moment and the torque data of the current moment is obtained to obtain the torque difference value. The torque difference value is normalized to obtain the normalized torque difference value. The product between the environmental stability and the normalized torque difference value is normalized to obtain the first bolt loosening confidence of any fastening bolt at the current moment.

[0020] Preferably, obtaining the second bolt loosening confidence level of any fastening bolt at the current moment based on the changing trend characteristics of the monitoring data in each dimension of the multidimensional monitoring data sequence includes:

[0021] In the multidimensional monitoring data sequence, multidimensional monitoring data that is the same as the wind speed data at the current moment is recorded as target multidimensional monitoring data. Among all target multidimensional monitoring data, vibration data from a preset number of target multidimensional monitoring data are selected to form a first vibration data subsequence. The vibration data at the current moment is added to the first vibration data subsequence to obtain a second vibration data subsequence.

[0022] The average value of the first vibration data subsequence is obtained and recorded as the historical vibration mean. The absolute value of the difference between the current vibration data and the historical vibration mean is normalized to obtain the first vibration deviation value. The standard deviations of the first vibration data subsequence and the second vibration subsequence are obtained respectively. The absolute value of the difference between the standard deviations of the first vibration data subsequence and the standard deviations of the second vibration subsequence is normalized to obtain the second vibration deviation value. The sum of the first vibration deviation value and the second vibration deviation value is obtained to obtain the vibration deviation value.

[0023] The temperature data in the multidimensional monitoring data sequence is fitted to obtain a first-order temperature fitting function. The inverse of the slope of the first-order temperature fitting function is normalized to obtain the temperature change value.

[0024] A temperature change graph is constructed by acquiring temperature data from a preset historical period prior to the current moment. The horizontal axis of the temperature change graph represents time, and the vertical axis represents temperature data. The last peak temperature data and the last valley temperature data are acquired from the temperature change graph. The difference between the last peak temperature data and the last valley temperature data is calculated to obtain the temperature difference value. A first interval time between the last peak temperature data and the last valley temperature data is acquired. The product of the reciprocal of the temperature difference and the first interval time is normalized to obtain the degree of temperature fluctuation. A second interval time between the last valley temperature data and the temperature data at the current moment is acquired. The sum of the degree of temperature fluctuation and the second interval time is calculated to obtain the torque confidence level.

[0025] The product of the vibration deviation value, the temperature change value, and the torque confidence level is normalized to obtain the second bolt loosening confidence level of any fastening bolt at the current moment.

[0026] Preferably, obtaining the aging degree of any of the fastening bolts includes:

[0027] The service life of any fastening bolt is obtained, and the service life of any fastening bolt is normalized to obtain the aging degree of any fastening bolt.

[0028] Preferably, obtaining the looseness degree of any fastening bolt at the current moment based on the suspected looseness degree of the bolt, the bolt looseness confidence level, and the aging degree of any fastening bolt includes:

[0029] The product of the preset aging influence coefficient, the aging degree of any fastening bolt, and the suspected loosening degree of the bolt is obtained to get the aging influence degree. The difference between the suspected loosening degree of the bolt and the aging influence degree is obtained to get the actual loosening degree.

[0030] The product of the bolt loosening confidence level and the actual loosening degree is obtained to determine the bolt loosening degree of any fastening bolt at the current moment.

[0031] Preferably, the step of adjusting the preset warning threshold based on the degree of looseness of any fastening bolt at the current moment to obtain an adaptive warning threshold for any fastening bolt at the current moment includes:

[0032] The threshold adjustment value is obtained by multiplying the preset threshold adjustment coefficient, the bolt loosening degree of any fastening bolt at the current moment, and the preset warning threshold. The difference between the preset warning threshold and the threshold adjustment value is obtained to obtain the adaptive warning threshold of any fastening bolt at the current moment.

[0033] Preferably, the step of providing a loosening warning for any of the fastening bolts based on the difference between the vibration data at the current moment and its adaptive warning threshold includes:

[0034] Based on the difference between the vibration data at the current moment and its adaptive warning threshold, it is determined whether any of the fastening bolts has become loose at the current moment;

[0035] An adaptive warning threshold is obtained by acquiring the vibration data of any fastening bolt at each moment. If any fastening bolt shows abnormal loosening at each moment within a continuous preset time, a loosening warning is issued for any fastening bolt.

[0036] Preferably, determining whether any fastening bolt exhibits abnormal loosening at the current moment based on the difference between the vibration data at the current moment and its adaptive warning threshold includes:

[0037] If the vibration data at the current moment is greater than or equal to the adaptive warning threshold of the vibration data at the current moment, then it is confirmed that any of the fastening bolts has become loose at the current moment.

[0038] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows:

[0039] In this invention, the suspected looseness of the bolts is first obtained, and the fluctuation characteristics of the vibration data are used to preliminarily determine whether the fastening bolts are loose. Then, the bolt looseness confidence is obtained, and multi-dimensional features are combined to reduce the influence of the environment on the vibration data. Next, the bolt looseness is obtained by combining the aging degree of the fastening bolts, which reduces the influence of bolt aging on the vibration data. Finally, the preset warning threshold is adjusted to obtain the adaptive warning threshold of the fastening bolts at the current moment, so that the adaptive warning threshold conforms to the actual operating conditions of the wind turbine and improves the accuracy of the warning. Attached Figure Description

[0040] 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.

[0041] Figure 1 This is a flowchart of a method for early warning of loosening of wind turbine tower fastening bolts provided in Embodiment 1 of the present invention. Detailed Implementation

[0042] 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.

[0043] 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.

[0044] To illustrate the technical solution of the present invention, specific embodiments are described below.

[0045] See Figure 1 This is a flowchart of a method for early warning of loosening of wind turbine tower fastening bolts provided in Embodiment 1 of the present invention. Figure 1 As shown, the method may include:

[0046] Step S101: For any fastening bolt in the wind turbine tower, according to a preset monitoring frequency, acquire multidimensional monitoring data of the fastening bolt at each moment to obtain a multidimensional monitoring data sequence within a preset time period including the current moment. The multidimensional monitoring data includes vibration data, wind speed data, temperature data, and torque data.

[0047] As a crucial component of wind turbines, the wind turbine tower bears the weight and wind load of the turbine, ensuring its stability during long-term operation. However, during long-term operation, the wind turbine tower is subjected to complex wind loads, temperature changes, vibrations, and environmental factors (such as humidity and salt spray), which may cause the fastening bolts to loosen. This loosening not only affects the structural stability of the wind turbine tower but may even lead to serious safety accidents. Therefore, it is essential to use relevant data on the fastening bolts of the wind turbine tower for early warning of loosening.

[0048] This embodiment uses sensors installed on the wind turbine tower to collect data related to the fastening bolts, providing early warning of bolt loosening. The data includes, but is not limited to, vibration data, wind speed data, temperature data, and torque data. Vibration sensors are installed on the tower wall around the flange connections at various sections; wind speed sensors are installed on the anemometer bracket at the top of the nacelle or at the front of the hub; temperature sensors are installed on the bolt surfaces at the flange connections; and torque sensors are installed in the strain monitoring area of ​​the bolt shank. Since vibration sensors are typically installed in multiple locations, the vibration data used in this embodiment is the average of data from multiple sensors.

[0049] Since the method for early warning of loosening of each fastening bolt on the wind turbine tower is the same, this embodiment uses a preset monitoring frequency of once per second for any fastening bolt in the wind turbine tower. This frequency is not limited here and can be set according to the specific implementation scenario. Vibration data, wind speed data, temperature data, and torque data of any fastening bolt at each moment are acquired to form multi-dimensional monitoring data. This results in a multi-dimensional monitoring data sequence within a preset time period including the current moment, used for early warning analysis of loosening of any fastening bolt in the wind turbine tower. In order to ensure that the multi-dimensional monitoring data sequence reflects the data changes of the fastening bolt, the preset time period in this embodiment is set to 5 minutes, which is not limited here and can be set according to the specific implementation scenario.

[0050] Traditional methods for early warning of loosening wind turbine tower bolts typically employ vibration signal analysis techniques with fixed thresholds. For example, they monitor the root mean square (RMS) value of the vibration signal and compare it to a preset threshold to determine if the tower bolts are loose, then issue an early warning. However, wind turbine operating conditions are complex (e.g., wind speed fluctuations, temperature changes, equipment aging), and environmental changes can cause normal vibration amplitudes to exceed the fixed threshold, resulting in a high false alarm rate. Furthermore, when bolts loosen slowly, the vibration amplitude change may not reach the fixed threshold, leading to a high risk of missed alarms. Additionally, traditional methods cannot adapt to equipment aging or load changes throughout the wind turbine's lifespan, making early warnings of loosening tower bolts inaccurate under fixed thresholds.

[0051] Therefore, this embodiment first obtains the suspected looseness of any fastening bolt at the current moment, then combines multi-dimensional features to obtain the bolt looseness confidence of any fastening bolt at the current moment, then combines the aging degree of any fastening bolt to obtain the bolt looseness degree, and finally adjusts the preset warning threshold to obtain the adaptive warning threshold of the fastening bolt at the current moment, so that the adaptive warning threshold conforms to the actual operating conditions of the wind turbine and improves the accuracy of the warning.

[0052] Step S102: Based on the fluctuation characteristics of the vibration data in the multidimensional monitoring data sequence, obtain the suspected looseness of any fastening bolt at the current moment; based on the data difference between the multidimensional monitoring data at the current moment and the multidimensional monitoring data at the previous moment, and the changing trend characteristics of the monitoring data in each dimension of the multidimensional monitoring data sequence, obtain the bolt loosening confidence of any fastening bolt at the current moment.

[0053] Since loosening of fastening bolts will lead to a decrease in the stiffness of the tower structure, the most obvious feature when the bolts are loose is that the vibration data collected by the vibration sensor installed on the wind turbine tower will show a significant change, that is, the vibration data will rise slowly or change dramatically in a transient manner. This corresponds to the early slight loosening of the fastening bolts and the sudden slippage and loosening of the fastening bolts. Therefore, the degree of suspected loosening of any fastening bolt at the current moment is obtained based on the fluctuation characteristics of the vibration data in the multidimensional monitoring data sequence.

[0054] The method for obtaining the suspected looseness of any fastening bolt at the current moment based on the fluctuation characteristics of vibration data in the multidimensional monitoring data sequence is as follows:

[0055] The absolute value of the difference between the vibration data at the current moment and the vibration data at the previous moment is obtained to get the instantaneous fluctuation value. The vibration data in the multidimensional monitoring data sequence is fitted using the least squares method to obtain a first-order vibration fitting function. The slope of the first-order vibration fitting function is obtained, and the slope is normalized by a sliding window to obtain the overall fluctuation value. The product of the instantaneous fluctuation value and the overall fluctuation value is obtained to get the suspected looseness of any fastening bolt at the current moment. The least squares method and sliding window normalization are existing technologies and will not be described in detail here.

[0056] In one embodiment, the formula for calculating the suspected looseness of any fastening bolt at the current moment is:

[0057]

[0058] Where A represents the suspected looseness of any fastening bolt at the current moment; This represents the vibration data of any fastening bolt at the current moment. This represents the vibration data of any fastening bolt at the previous moment in the current moment. The slope of the fitting function for a single vibration is normalized by a sliding window, which is also the overall fluctuation value. It is the absolute value symbol.

[0059] It should be noted that, This represents the instantaneous fluctuation value of the vibration data of any fastening bolt at the current moment. The larger A is, the more the vibration data of any fastening bolt matches the situation of sudden slippage and loosening; the larger K is, the more the vibration data of any fastening bolt matches the situation of early slight loosening.

[0060] Besides the effects of loose bolts on vibration data, increased wind speed and temperature also cause changes in vibration signals. Torque data is also closely related to bolt loosening, but it is also affected by temperature changes. Therefore, after obtaining the suspected looseness of any bolt at the current moment, it is necessary to combine multi-dimensional features to obtain the bolt looseness confidence level of any bolt at the current moment to reduce the influence of the environment on vibration data. The method for obtaining the bolt looseness confidence level of any bolt at the current moment is as follows:

[0061] (1) When the fastening bolts are loose, the wind speed data and temperature data will not change significantly, while the torque data will be smaller due to the loosening of the fastening bolts. Therefore, based on the data difference between the current multidimensional monitoring data and the previous multidimensional monitoring data, the first bolt loosening confidence of any fastening bolt at the current moment can be obtained.

[0062] Specifically, the absolute value of the difference between the wind speed data at the current moment and the wind speed data at the previous moment is obtained to get the wind speed difference value; the absolute value of the difference between the temperature data at the current moment and the temperature data at the previous moment is obtained to get the temperature difference value; the product of the wind speed difference value and the temperature difference value is obtained to get the environmental change value; and the reciprocal of the sum of the environmental change value and a preset constant is obtained to get the environmental stability level.

[0063] The difference between the torque data of the previous moment and the torque data of the current moment is obtained to obtain the torque difference value. The torque difference value is then subjected to sliding window normalization to obtain the normalized torque difference value. Sliding window normalization is an existing technology and will not be elaborated here. The product between the environmental stability and the normalized torque difference value is then normalized to obtain the first bolt loosening confidence of any fastening bolt at the current moment.

[0064] In one embodiment, the formula for calculating the confidence level of the first bolt loosening at the current moment for any fastening bolt is:

[0065]

[0066] Where B1 is the first bolt loosening confidence level of any fastening bolt at the current moment; For any fastening bolt, the wind speed data at the current moment; For any fastening bolt, the wind speed data from the previous moment at the current moment; The temperature data of any fastening bolt at the current moment; This represents the temperature data of any fastening bolt at the previous time point. The normalized torque difference value is the difference between the torque data of any fastening bolt at the current moment and the torque data at the previous moment, after being normalized by a sliding window. is the normalization function; c is a preset constant. In this embodiment, c=0.01 is set to ensure that the fraction is meaningful. There is no restriction here, and it can be set according to the specific implementation scenario.

[0067] It should be noted that, This represents the wind speed difference value. This represents the temperature difference value. or The larger the value, the greater the change in wind speed and temperature data at the current moment, and the less consistent it is with the changes in wind speed and temperature caused by loosening of fastening bolts. The smaller it is; This is the normalized torque difference value. The larger B1 is, the more closely the change in torque data at the current moment matches the torque change caused by the loosening of the fastening bolts.

[0068] (2) Based on the changing trend characteristics of the monitoring data in each dimension of the multidimensional monitoring data sequence, obtain the second bolt loosening confidence of any fastening bolt at the current moment.

[0069] When the fastening bolts loosen, the local characteristics of each dimension of the monitoring data will also show related features: First, when the wind speed increases, the load on the tower increases. Under normal operating conditions, the vibration amplitude is positively correlated with the wind speed, but when the fastening bolts loosen, the vibration amplitude will be abnormally high at the same wind speed. Therefore, in the multidimensional monitoring data sequence, the multidimensional monitoring data that is the same as the wind speed data at the current moment is recorded as the target multidimensional monitoring data. In order to make the target multidimensional monitoring data closer to the current operating conditions, in this embodiment, all target multidimensional monitoring data are sorted according to time, and the vibration data in the last preset number of target multidimensional monitoring data are selected to form the first vibration data subsequence. In this embodiment, the preset number is set to 20, which is not limited here and can be set according to the specific implementation scenario. The vibration data at the current moment is added to the first vibration data subsequence to obtain the second vibration data subsequence.

[0070] The average value of the first vibration data subsequence is obtained and recorded as the historical vibration mean. The absolute value of the difference between the current vibration data and the historical vibration mean is normalized by a sliding window to obtain the first vibration deviation value. The standard deviations of the first vibration data subsequence and the second vibration subsequence are obtained respectively. The absolute value of the difference between the standard deviations of the first vibration data subsequence and the standard deviations of the second vibration subsequence is normalized by a sliding window to obtain the second vibration deviation value. The sum of the first vibration deviation value and the second vibration deviation value is obtained to obtain the vibration deviation value.

[0071] Second, under normal operating conditions, temperature data and vibration data are not related, but a decrease in temperature may cause the fastening bolts to shrink, resulting in an increase in vibration amplitude. Therefore, the least squares method is used to fit the temperature data in the multidimensional monitoring data sequence to obtain a first-order temperature fitting function. The least squares method is an existing technology and will not be described in detail here. The inverse of the slope of the first-order temperature fitting function is normalized by a sliding window to obtain the temperature change value.

[0072] Third, loosening of fasteners directly manifests as a decrease in torque data, and the rate of decrease is positively correlated with the degree of loosening. Temperature changes also cause torque changes, but these changes are delayed. Therefore, a temperature change graph is constructed by acquiring temperature data from a preset historical period prior to the current moment. The horizontal axis of the temperature change graph represents time, and the vertical axis represents temperature data. The last peak temperature data and the last valley temperature data are obtained from the temperature change graph. The difference between the last peak temperature data and the last valley temperature data is calculated to obtain the temperature difference value. The first interval time between the last peak temperature data and the last valley temperature data is obtained. The product of the reciprocal of the temperature difference value and the first interval time is subjected to sliding window normalization to obtain the degree of temperature fluctuation. The second interval time between the last valley temperature data and the temperature data at the current moment is obtained. The sum of the degree of temperature fluctuation and the second interval time is calculated to obtain the torque confidence level. The sliding window normalization is an existing technology and will not be elaborated here.

[0073] The product of the vibration deviation value, the temperature change value, and the torque confidence level is normalized to obtain the second bolt loosening confidence level of any fastening bolt at the current moment.

[0074] In one embodiment, the formula for calculating the confidence level of the second bolt loosening at the current moment for any fastening bolt is:

[0075]

[0076] Where B2 is the second bolt loosening confidence level of any fastening bolt at the current moment; The first vibration deviation value is the value after normalizing the absolute value of the difference between the current vibration data and the historical vibration mean using a sliding window. The second vibration deviation value is the value after sliding window normalization of the absolute value of the difference between the standard deviation of the first vibration data subsequence and the standard deviation of the second vibration subsequence. The temperature change value is the value obtained by normalizing the inverse of the slope of the first-order temperature fitting function using a sliding window. The degree of temperature fluctuation is the value obtained by normalizing the product of the reciprocal of the temperature difference and the first time interval using a sliding window. This is the second interval time; This is the normalization function.

[0077] It should be noted that, This represents the vibration deviation value. or The larger the value, the more the vibration data deviates from historical conditions and the greater the deviation from historical vibration data fluctuations at the current wind speed. The larger the value, the more it matches the deviation characteristics of vibration data when fastening bolts are loose. The larger it is; For the temperature change value, when the slope of the first-order temperature fitting function is negative, the larger the slope, the more significant the temperature drop. The smaller the value, the more significantly the vibration data is affected by temperature, and the less consistent it is with the characteristics of loosening fasteners. The smaller it is; For torque confidence level, The smaller the value, the smaller the intensity of the most recent temperature fluctuation. The larger the value, the longer the interval between the most recent temperature fluctuation and the current moment, which is less consistent with the characteristics of torque data changes caused by temperature variations. The larger it is, the more... The larger it is.

[0078] (3) Based on the average between the first bolt loosening confidence level and the second bolt loosening confidence level, the bolt loosening confidence level of any fastening bolt at the current moment is obtained.

[0079] In one embodiment, the formula for calculating the bolt loosening confidence level of any fastening bolt at the current moment is:

[0080]

[0081] Where B is the confidence level of bolt loosening of any fastening bolt at the current moment; B1 is the confidence level of first bolt loosening of any fastening bolt at the current moment; and B2 is the confidence level of second bolt loosening of any fastening bolt at the current moment.

[0082] Thus, the bolt loosening confidence level of any fastening bolt at the current moment is obtained.

[0083] Step S103: Obtain the aging degree of any fastening bolt. Based on the suspected loosening degree of the bolt, the loosening confidence of the bolt, and the aging degree of any fastening bolt, obtain the bolt loosening degree of any fastening bolt at the current moment.

[0084] After obtaining the bolt loosening confidence level of any fastening bolt at the current moment, since bolt aging will also cause changes in vibration data, that is, bolt aging will cause the vibration data fluctuation to increase normally, the service life of the fastening bolt is obtained. The service life of the fastening bolt is then subjected to maximum and minimum normalization to obtain the aging degree of the fastening bolt. Maximum and minimum normalization is a prior art, with the minimum being 0 and the maximum being the maximum service life of the fastening bolt, which will not be elaborated here. Further, based on the suspected bolt loosening degree, bolt loosening confidence level and the aging degree of any fastening bolt, the bolt loosening degree of any fastening bolt at the current moment is obtained.

[0085] The method for obtaining the bolt loosening degree of any fastening bolt at the current moment based on the suspected bolt loosening degree, bolt loosening confidence level, and the aging degree of any fastening bolt is as follows:

[0086] The product of the preset aging influence coefficient, the aging degree of any fastening bolt, and the suspected loosening degree of the bolt is obtained to get the aging influence degree. The difference between the suspected loosening degree of the bolt and the aging influence degree is obtained to get the actual loosening degree.

[0087] The product of the bolt loosening confidence level and the actual loosening degree is obtained to determine the bolt loosening degree of any fastening bolt at the current moment.

[0088] In one embodiment, the formula for calculating the degree of looseness of any fastening bolt at the current moment is:

[0089]

[0090] Where E represents the degree of looseness of any fastening bolt at the current moment; B represents the confidence level of bolt looseness of any fastening bolt at the current moment; A represents the suspected degree of looseness of any fastening bolt at the current moment; D represents the degree of aging of any fastening bolt; 0.1 is the preset aging influence coefficient in this embodiment. Since bolt aging is not a major factor, the preset aging influence coefficient is set to 0.1 in this embodiment. This is not a limitation and can be set according to the specific implementation scenario.

[0091] It should be noted that the larger A is, the greater the probability that any fastening bolt will loosen at the current moment, and the larger E is; the larger B is, the greater the credibility that any fastening bolt will actually loosen at the current moment, and the larger E is; since bolt aging will cause the vibration data fluctuation to increase normally, the larger D is, the greater the vibration data fluctuation of any fastening bolt, and the more normal the increase in the vibration data fluctuation of any fastening bolt is, the smaller E is.

[0092] Thus, the degree of looseness of any fastening bolt at the current moment is obtained.

[0093] Step S104: Adjust the preset warning threshold according to the degree of looseness of any fastening bolt at the current moment to obtain the adaptive warning threshold of any fastening bolt at the current moment, and issue a loosening warning for any fastening bolt based on the difference between the vibration data at the current moment and the adaptive warning threshold.

[0094] After obtaining the degree of looseness of any fastening bolt at the current moment, the preset warning threshold is adjusted based on this degree of looseness to obtain an adaptive warning threshold for any fastening bolt at the current moment. The method for obtaining the adaptive warning threshold for any fastening bolt at the current moment is as follows:

[0095] The threshold adjustment value is obtained by multiplying the preset threshold adjustment coefficient, the bolt loosening degree of any fastening bolt at the current moment, and the preset warning threshold. The difference between the preset warning threshold and the threshold adjustment value is obtained to obtain the adaptive warning threshold of any fastening bolt at the current moment.

[0096] In one embodiment, the formula for calculating the adaptive warning threshold of any fastening bolt at the current moment is:

[0097]

[0098] in, The adaptive warning threshold for any fastening bolt at the current moment; The preset warning threshold is generally given by relevant experts after evaluating the current model of wind turbine tower. There is no restriction here, and it can be set according to the specific implementation scenario. E is the degree of looseness of any fastening bolt at the current moment. 0.5 is the preset threshold adjustment coefficient to avoid the threshold being too small. The maximum adjustment range of the warning threshold should be half of the original threshold. Therefore, in this embodiment, 0.5 is set as the preset threshold adjustment coefficient. There is no restriction here, and it can be set according to the specific implementation scenario.

[0099] It should be noted that the larger the value of E, the greater the looseness of any fastening bolt. Due to external environmental factors, the vibration data of any fastening bolt at the current moment may not reach the preset warning threshold. Therefore, it is necessary to reduce the preset warning threshold. The smaller it is.

[0100] Furthermore, after obtaining the adaptive warning threshold for any fastening bolt at the current moment, if the vibration data at the current moment is greater than or equal to the adaptive warning threshold, then it is confirmed that any fastening bolt has become loose at the current moment. To prevent false alarms caused by communication interference, the adaptive warning threshold for the vibration data of any fastening bolt at each moment is obtained. If any fastening bolt becomes loose at every moment within a consecutive preset time period, a loosening warning is issued for any fastening bolt. Relevant personnel are then notified to inspect the current wind turbine tower fastening bolts, investigate the cause of the loosening, and repair them promptly. In this embodiment, the preset time is set to 5 seconds. The preset time needs to be set according to the user's actual needs and is not limited here; it can be set according to the specific implementation scenario.

[0101] In summary, the embodiments of the present invention first obtain the suspected looseness of the bolts, and preliminarily determine whether the bolts are loose based on the fluctuation characteristics of the vibration data; then, obtain the bolt looseness confidence level, and combine multi-dimensional features to reduce the influence of the environment on the vibration data; then, combine the aging degree of the fastening bolts to obtain the bolt looseness degree, and reduce the influence of bolt aging on the vibration data; finally, adjust the preset warning threshold to obtain the adaptive warning threshold of the fastening bolts at the current moment, so that the adaptive warning threshold conforms to the actual operating conditions of the wind turbine and improves the accuracy of the warning.

[0102] 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. A method for early warning of loosening of wind turbine tower fastening bolts, characterized in that, The aforementioned method for early warning of loosening of wind turbine tower fastening bolts includes: For any fastening bolt in the wind turbine tower, according to a preset monitoring frequency, multidimensional monitoring data of the fastening bolt at each moment is acquired to obtain a multidimensional monitoring data sequence within a preset time period including the current moment. The multidimensional monitoring data includes vibration data, wind speed data, temperature data, and torque data. Based on the fluctuation characteristics of the vibration data in the multidimensional monitoring data sequence, the suspected looseness of any fastening bolt at the current moment is obtained. Based on the data difference between the multidimensional monitoring data at the current moment and the multidimensional monitoring data at the previous moment, as well as the changing trend characteristics of the monitoring data in each dimension of the multidimensional monitoring data sequence, the bolt looseness confidence of any fastening bolt at the current moment is obtained. Obtain the aging degree of any of the fastening bolts, and based on the suspected loosening degree of the bolt, the loosening confidence of the bolt, and the aging degree of any of the fastening bolts, obtain the bolt loosening degree of any of the fastening bolts at the current moment; Based on the degree of looseness of any fastening bolt at the current moment, the preset warning threshold is adjusted to obtain the adaptive warning threshold of any fastening bolt at the current moment. Based on the difference between the vibration data at the current moment and the adaptive warning threshold, a loosening warning is issued for any fastening bolt. The step of obtaining the suspected looseness of any fastening bolt at the current moment based on the fluctuation characteristics of the vibration data in the multidimensional monitoring data sequence includes: The absolute value of the difference between the vibration data at the current moment and the vibration data at the previous moment is obtained to obtain the instantaneous fluctuation value. The vibration data in the multidimensional monitoring data sequence is fitted to obtain a first-order vibration fitting function. The slope of the first-order vibration fitting function is obtained. The slope is normalized to obtain the overall fluctuation value. The product of the instantaneous fluctuation value and the overall fluctuation value is obtained to obtain the suspected looseness of any fastening bolt at the current moment. The step of obtaining the looseness degree of any fastening bolt at the current moment based on the suspected looseness degree of the bolt, the bolt looseness confidence level, and the aging degree of any fastening bolt includes: The product of the preset aging influence coefficient, the aging degree of any fastening bolt, and the suspected loosening degree of the bolt is obtained to get the aging influence degree. The difference between the suspected loosening degree of the bolt and the aging influence degree is obtained to get the actual loosening degree. The product of the bolt loosening confidence level and the actual loosening degree is obtained to determine the bolt loosening degree of any fastening bolt at the current moment.

2. The method for early warning of loosening of wind turbine tower fastening bolts according to claim 1, characterized in that, The step of obtaining the bolt loosening confidence level of any fastening bolt at the current moment based on the data difference between the current multidimensional monitoring data and the previous multidimensional monitoring data, and the changing trend characteristics of the monitoring data in each dimension of the multidimensional monitoring data sequence, includes: Based on the data difference between the current multidimensional monitoring data and the previous multidimensional monitoring data, the first bolt loosening confidence level of any fastening bolt at the current moment is obtained; Based on the changing trend characteristics of the monitoring data in each dimension of the multidimensional monitoring data sequence, the second bolt loosening confidence level of any fastening bolt at the current moment is obtained; The bolt loosening confidence level of any fastening bolt at the current moment is obtained by averaging the first bolt loosening confidence level and the second bolt loosening confidence level.

3. The method for early warning of loosening of wind turbine tower fastening bolts according to claim 2, characterized in that, The step of obtaining the first bolt loosening confidence level of any fastening bolt at the current moment based on the data difference between the multidimensional monitoring data at the current moment and the multidimensional monitoring data at the previous moment includes: Obtain the absolute value of the difference between the wind speed data at the current moment and the wind speed data at the previous moment to obtain the wind speed difference value; obtain the absolute value of the difference between the temperature data at the current moment and the temperature data at the previous moment to obtain the temperature difference value; obtain the product of the wind speed difference value and the temperature difference value to obtain the environmental change value; obtain the reciprocal of the sum of the environmental change value and the preset constant to obtain the environmental stability level. The difference between the torque data of the previous moment and the torque data of the current moment is obtained to obtain the torque difference value. The torque difference value is normalized to obtain the normalized torque difference value. The product between the environmental stability and the normalized torque difference value is normalized to obtain the first bolt loosening confidence of any fastening bolt at the current moment.

4. The method for early warning of loosening of wind turbine tower fastening bolts according to claim 2, characterized in that, The step of obtaining the second bolt loosening confidence level of any fastening bolt at the current moment based on the changing trend characteristics of the monitoring data in each dimension of the multidimensional monitoring data sequence includes: In the multidimensional monitoring data sequence, multidimensional monitoring data that is the same as the wind speed data at the current moment is recorded as target multidimensional monitoring data. Among all target multidimensional monitoring data, vibration data from a preset number of target multidimensional monitoring data are selected to form a first vibration data subsequence. The vibration data at the current moment is added to the first vibration data subsequence to obtain a second vibration data subsequence. The average value of the first vibration data subsequence is obtained and recorded as the historical vibration mean. The absolute value of the difference between the current vibration data and the historical vibration mean is normalized to obtain the first vibration deviation value. The standard deviations of the first vibration data subsequence and the second vibration subsequence are obtained respectively. The absolute value of the difference between the standard deviations of the first vibration data subsequence and the standard deviations of the second vibration subsequence is normalized to obtain the second vibration deviation value. The sum of the first vibration deviation value and the second vibration deviation value is obtained to obtain the vibration deviation value. The temperature data in the multidimensional monitoring data sequence is fitted to obtain a first-order temperature fitting function. The inverse of the slope of the first-order temperature fitting function is normalized to obtain the temperature change value. A temperature change graph is constructed by acquiring temperature data from a preset historical period prior to the current moment. The horizontal axis of the temperature change graph represents time, and the vertical axis represents temperature data. The last peak temperature data and the last valley temperature data are acquired from the temperature change graph. The difference between the last peak temperature data and the last valley temperature data is calculated to obtain the temperature difference value. A first interval time between the last peak temperature data and the last valley temperature data is acquired. The product of the reciprocal of the temperature difference and the first interval time is normalized to obtain the degree of temperature fluctuation. A second interval time between the last valley temperature data and the temperature data at the current moment is acquired. The sum of the degree of temperature fluctuation and the second interval time is calculated to obtain the torque confidence level. The product of the vibration deviation value, the temperature change value, and the torque confidence level is normalized to obtain the second bolt loosening confidence level of any fastening bolt at the current moment.

5. The method for early warning of loosening of wind turbine tower fastening bolts according to claim 1, characterized in that, The process of obtaining the aging degree of any of the fastening bolts includes: The service life of any fastening bolt is obtained, and the service life of any fastening bolt is normalized to obtain the aging degree of any fastening bolt.

6. The method for early warning of loosening of wind turbine tower fastening bolts according to claim 1, characterized in that, The step of adjusting the preset warning threshold based on the degree of looseness of any fastening bolt at the current moment to obtain an adaptive warning threshold for any fastening bolt at the current moment includes: The threshold adjustment value is obtained by multiplying the preset threshold adjustment coefficient, the bolt loosening degree of any fastening bolt at the current moment, and the preset warning threshold. The difference between the preset warning threshold and the threshold adjustment value is obtained to obtain the adaptive warning threshold of any fastening bolt at the current moment.

7. The method for early warning of loosening of wind turbine tower fastening bolts according to claim 1, characterized in that, The step of providing a loosening warning for any of the fastening bolts based on the difference between the current vibration data and its adaptive warning threshold includes: Based on the difference between the vibration data at the current moment and its adaptive warning threshold, it is determined whether any of the fastening bolts has become loose at the current moment; An adaptive warning threshold is obtained by acquiring the vibration data of any fastening bolt at each moment. If any fastening bolt shows abnormal loosening at each moment within a continuous preset time, a loosening warning is issued for any fastening bolt.

8. The method for early warning of loosening of wind turbine tower fastening bolts according to claim 7, characterized in that, The step of determining whether any fastening bolt is abnormally loose at the current moment based on the difference between the vibration data at the current moment and its adaptive warning threshold includes: If the vibration data at the current moment is greater than or equal to the adaptive warning threshold of the vibration data at the current moment, then it is confirmed that any of the fastening bolts has become loose at the current moment.

Images