Extreme weather power generation equipment fault prediction method based on time sequence large model

Abstract

The invention relates to the technical field of generator fault prediction, in particular to an extreme weather power generation equipment fault prediction method based on a time sequence large model, and the method comprises the steps: synchronously collecting multiple parameters; judging an abnormal event; identifying trend types; risk level assessment; adjusting a dynamic threshold value; and risk sequence prediction and early warning. A plurality of parameters which are closely related to structural load, electrical disturbance and thermal-lubrication states are introduced through the three-axis vibration peak value in the middle of a blade, the oil viscosity of a gearbox, the current harmonic wave of a stator, the starting torque increment of a yaw system, the wind speed change rate of the top of a cabin and the temperature rise rate of a main shaft bearing; full-chain monitoring of the operation state of the wind turbine generator in the strong wind extreme weather is achieved, and the problems that due to incomplete monitoring parameter coverage and fixed threshold value setting, lubrication abnormity trend recognition lags behind in the extreme weather, fault early warning is not timely, and the unit shutdown loss is aggravated are effectively solved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of generator fault prediction, in particular to an extreme weather power generation equipment fault prediction method based on a time sequence large model. BACKGROUND

[0002] With the continuous growth of wind power installed capacity, the operation stability of wind power generation equipment under extreme weather conditions has become increasingly prominent. In particular, during strong winds, typhoons and other extreme weather processes, wind turbines are affected by aerodynamic loads, structural responses and electrical disturbances, and are prone to induce lubrication degradation, component fatigue and system instability.

[0003] Traditional wind power operation and maintenance methods are mostly based on single-point monitoring or fixed threshold judgment, which cannot dynamically perceive the coordinated change trend of multiple parameters, and cannot meet the demand for early fault trend identification and early warning under extreme environments. In recent years, the development of time sequence prediction models has provided a new means for time domain evolution analysis of multi-source monitoring data, but it is still necessary to build a monitoring parameter system that matches the actual unit structure response and introduce a dynamic adjustment mechanism to adapt to the strong nonlinearity and rapid evolution of extreme wind operation scenarios. Therefore, it is urgent to propose a power generation equipment fault prediction method that can integrate key structural feature parameters, dynamic judgment mechanism and high-precision time sequence prediction capability to realize early identification and risk warning of fault trends under extreme weather conditions. SUMMARY

[0004] Therefore, the application provides an extreme weather power generation equipment fault prediction method based on a time sequence large model to overcome the problem that the existing technology causes the fault warning to be not timely and the unit shutdown loss to be aggravated due to the incomplete monitoring parameter coverage and the fixed threshold setting, which leads to the lag of lubrication abnormal trend identification under extreme weather.

[0005] To achieve the above purpose, the application provides an extreme weather power generation equipment fault prediction method based on a time sequence large model, comprising:

[0006] Real-time synchronous acquisition of three-axis vibration peak values of a blade middle portion of each wind turbine, oil viscosity of a gear box input shaft, current harmonic content of a generator stator, yaw system startup torque increment, wind speed variation rate of a cabin top portion and temperature rise rate of a main shaft bearing during strong wind passage;

[0007] According to the three-axis vibration peak values in a preset abnormal judgment period and a preset gust impact threshold, it is determined that an abnormal event exists, and an abnormal judgment result is obtained;

[0008] Based on the abnormal judgment result, according to the oil viscosity change rate and the temperature rise rate, it is determined that the type of the abnormal event is a lubrication abnormal trend, and a type judgment result is obtained;

[0009] Based on the type determination result, according to each of the current harmonic content, the three-axis vibration peak value and the preset risk evolution threshold, the risk level of the fault trend is determined as a serious warning, and a risk determination result is obtained;

[0010] Based on the risk determination result, the preset gust impact threshold is adjusted according to the starting torque increment and the wind speed change rate in the next preset first adjustment period, and the risk evolution threshold is adjusted according to the number of times of adjusting the gust impact threshold and the oil viscosity in the next preset second adjustment period;

[0011] According to all the risk determination results obtained after adjusting the risk evolution threshold in a preset prediction period, the three-axis vibration peak value and a preset time sequence model, a risk prediction sequence is generated, and a warning alarm is sent according to the time stamp of the risk prediction sequence.

[0012] Further, according to the oil viscosity change rate and the temperature rise rate, the type of the abnormal event is determined as a lubrication abnormal trend, and the process of obtaining a type determination result includes:

[0013] According to the oil viscosity change rate and the temperature rise rate, the change correlation amplitude is determined in a preset abnormal trend determination period;

[0014] According to the comparison result of the change correlation amplitude and the preset correlation amplitude threshold, the type of the abnormal event is determined as a lubrication abnormal trend, and the type determination result is obtained.

[0015] Further, according to each of the current harmonic content, the three-axis vibration peak value and the preset risk evolution threshold of each wind turbine, the risk level of the fault trend is determined as a serious warning, and the process includes:

[0016] According to the current harmonic content from the initial time to each time in a preset level determination period, a harmonic content fluctuation set is determined;

[0017] According to the harmonic content fluctuation set of each of any two adjacent wind turbines, a number of concerned wind turbines are determined from the wind turbine generator unit;

[0018] According to the three-axis vibration peak value and the preset risk evolution threshold of each of any two adjacent concerned wind turbines, the risk level of the fault trend is determined as a serious warning.

[0019] Further, according to the harmonic content fluctuation set of each of any two adjacent wind turbines, a number of concerned wind turbines are determined from the wind turbine generator unit, and the process includes:

[0020] According to two harmonic content fluctuation sets, a fluctuation consistency degree is determined;

[0021] Based on the comparison between the fluctuation consistency and the preset consistency threshold, several generators of interest are identified from the wind turbine units.

[0022] Furthermore, the process of determining the risk level of the fault trend as a severe warning based on the triaxial vibration peak values ​​of any two adjacent generators of interest and a preset risk evolution threshold includes:

[0023] Several peak value variations are determined based on the triaxial vibration peak values ​​of any two adjacent generators of interest;

[0024] The peak change fluctuation value is determined based on all the aforementioned peak change values;

[0025] Based on the comparison between the peak change fluctuation value and the preset risk evolution threshold, the risk level of the fault trend is determined to be a severe warning.

[0026] Furthermore, the process of adjusting the preset gust impact threshold based on the increment of the starting torque and the rate of change of wind speed within the next preset first adjustment cycle includes:

[0027] The synchronization degree is determined based on all the aforementioned starting torque increments and all the aforementioned wind speed change rates;

[0028] The preset gust impact threshold is adjusted based on the comparison between the adjusted synchronization degree and the preset adjusted synchronization threshold.

[0029] Furthermore, the process of adjusting the risk evolution threshold based on the number of times the gust impact threshold is adjusted within the next preset second adjustment cycle and the oil viscosity includes:

[0030] The fluctuation value of the number of adjustments is determined based on the total number of times the gust impact threshold is adjusted, and the fluctuation value of the oil viscosity is determined based on the total oil viscosity.

[0031] The risk evolution threshold is adjusted based on the fluctuation value of the number of adjustments and the fluctuation value of the oil viscosity.

[0032] Furthermore, the process of adjusting the risk evolution threshold based on the fluctuation value of the adjustment frequency and the fluctuation value of the oil viscosity includes:

[0033] The fluctuation deviation is determined based on the fluctuation value of the adjustment number and the preset fluctuation threshold, and the fluctuation is also determined based on the oil viscosity fluctuation.

[0034] The viscosity fluctuation deviation is determined by the value and the preset viscosity fluctuation threshold;

[0035] The risk evolution threshold is adjusted based on the comparison results of the frequency fluctuation deviation and the preset frequency fluctuation deviation threshold, and the comparison results of the viscosity fluctuation deviation and the preset viscosity fluctuation deviation threshold.

[0036] Furthermore, the process of issuing early warning alerts based on the timestamps of the risk prediction sequence includes:

[0037] Determine the degree of time distribution based on all timestamps;

[0038] Based on the comparison between the time distribution degree and the preset distribution degree threshold, it is determined that the wind turbine unit has malfunctioned, and the early warning alarm is issued.

[0039] Furthermore, the process of determining the existence of an abnormal event based on the triaxial vibration peak value and the preset gust impact threshold within the preset anomaly determination period, and obtaining the anomaly determination result, includes:

[0040] The degree of abnormal distribution is determined based on the comparison between the triaxial vibration peak value and the preset vibration peak value threshold.

[0041] The abnormal event is determined to exist based on the comparison between the abnormal distribution degree and the preset gust impact threshold, and the abnormality determination result is obtained.

[0042] Compared with existing technologies, the beneficial effects of this invention lie in the fact that by introducing multiple parameters closely related to structural loads, electrical disturbances, and thermal-lubrication conditions—such as the triaxial vibration peak value in the middle of the blade, the oil viscosity of the gearbox, the current harmonics of the stator, the starting torque increment of the yaw system, the wind speed change rate at the top of the nacelle, and the temperature rise rate of the main shaft bearing—it achieves full-chain monitoring of the wind turbine's operating status under extreme strong wind weather. Gust impacts directly cause intensified blade vibration, which is transmitted to the gearbox, causing fluctuations in oil viscosity and changes in the temperature rise rate. Lubrication degradation further intensifies component friction, leading to… Reflected in the changes in yaw torque and stator current harmonics; the risk level assessment integrates the coupling trend of electrical and mechanical signals, thereby driving dynamic threshold adjustment, realizing adaptive optimization of the model, and finally generating a continuous fault risk prediction sequence through a time-series large model. It can accurately judge the trend and provide early warning under multi-source coupled disturbance conditions, improve the fault perception sensitivity and operational safety of wind power equipment, and effectively solve the problem that due to incomplete monitoring parameter coverage and fixed threshold settings, the identification of abnormal lubrication trends under extreme weather conditions is delayed, resulting in untimely fault warnings and increased unit downtime losses.

[0043] Furthermore, by normalizing the rate of change in oil viscosity and the rate of temperature rise in the spindle bearing (existing technology, not elaborated further), the two types of data are analyzed on a unified scale, thereby improving the accuracy of measuring the consistency of their trend changes. Using the Pearson correlation coefficient as a measure of the strength of the correlation helps to accurately determine whether the deterioration of lubrication performance and the temperature rise have a synchronous trend. When the two are highly positively correlated, it indicates that the decline in lubrication performance has significantly affected heat conduction and mechanical friction, exhibiting the typical characteristic of lubrication degradation leading to increased temperature rise. This method effectively reveals the linkage evolution law between lubrication state and thermal effect during the formation of abnormal lubrication trends, thereby enhancing the accuracy and explanatory power of type determination.

[0044] Furthermore, by first calculating the standard deviation of the current harmonic content of each wind turbine within the judgment period, the degree of dynamic load instability caused by wind shear or control fluctuations on the electrical side is quantified. Then, the electrical fluctuation sets of adjacent wind turbines are compared to screen out "units of concern" with abnormal harmonic synchronization, so as to lock down equipment that may be affected by the same local extreme wind field. Finally, by combining the triaxial vibration peak value of these units with the preset risk evolution threshold, the units most prone to failure under electromechanical dual stress can be accurately identified, avoiding missed judgments caused by simply relying on the overall average index, and issuing the most targeted severe warning in advance.

[0045] Furthermore, by calculating the correlation coefficient of harmonic fluctuations in the current of adjacent wind turbines, units exhibiting synchronous electrical responses under the same wind field disturbances can be identified. Since simultaneous harmonic amplitude fluctuations in neighboring units indicate they may be affected by similar gusts or load switching, these units are more prone to cascading effects on their structures and lubrication systems. By designating units with fluctuation consistency exceeding a threshold as "generators of concern," subsequent peak vibration analysis and risk assessment can focus on key areas truly impacted by extreme weather. This reduces the overall computational load and avoids misjudging local disturbances as overall field risks, significantly improving the accuracy and efficiency of fault warnings and responses.

[0046] Furthermore, by focusing on the differences in the peak values ​​of triaxial vibrations between generators, fluctuation characteristics representing the severity of structural response are extracted and their overall instability is quantified in the form of standard deviation. This is then compared with a preset risk evolution threshold. When the fluctuation exceeds the threshold, a severe warning level is triggered, reflecting the dynamic coordination relationship of the mechanical states between units. When the degree of disturbance of adjacent units under extreme weather conditions deviates significantly, it can be regarded as a direct signal of increased possibility of fault evolution, thereby realizing the early identification of fault trends and the judgment of risk escalation.

[0047] Furthermore, by introducing synchronization degree and cosine similarity as indicators to measure the dynamic coupling relationship between the rate of change of wind speed and the fluctuation of the increment of starting torque, this embodiment can effectively identify the consistency of the yaw system's response to wind speed disturbances. Then, based on the relative deviation between the two, a set threshold adjustment coefficient is used to quantify the gain of the preset gust impact threshold. This not only avoids misjudging short-term synchronization fluctuations as anomalies but also achieves dynamic adjustment of the threshold, enabling the system to have adaptive adjustment capabilities. This method integrates temporal volatility and trend consistency at the numerical level, improving the ability to reasonably judge the operating boundaries of wind turbines under complex weather conditions, thereby enhancing the robustness of anomaly identification and the accuracy of the early warning threshold.

[0048] Furthermore, by introducing a strategy of co-correcting the risk evolution threshold using the fluctuation values ​​of adjustment frequency and oil viscosity, the operational stability and lubrication status changes of wind turbine units during strong winds can be dynamically reflected. Frequent adjustments to the gust impact threshold, coupled with increased volatility, often indicate severe external wind field disturbances or abnormal control system response. Simultaneously, oil viscosity fluctuations reflect the stress and thermal state changes in the gearbox lubrication system. Using the standard deviation of both as a quantitative basis to adjust the risk evolution threshold not only enhances the system's sensitivity to potential fault risks but also effectively filters out occasional disturbances, improving the stability and robustness of the early warning mechanism. By establishing a linkage mechanism between the control system's adjustment frequency and key mechanical performance indicators, adaptive optimization of the fault risk trend judgment threshold can be achieved.

[0049] Furthermore, by comprehensively analyzing the relative changes in the frequency of adjustment fluctuations and the viscosity fluctuations of the oil, the risk evolution threshold is reasonably increased to dynamically adapt to the operating status of wind turbines under extreme weather conditions. The frequency of adjustment reflects the frequency of gust impact threshold adjustments, while viscosity fluctuations reflect the stability of the lubrication system; both jointly affect the risk level of equipment failure development. By introducing a preset risk adjustment coefficient and corresponding weights, the two deviations are weighted and calculated to achieve precise adjustment of the risk threshold. This ensures that the fault early warning system can sensitively capture potential risks while avoiding misjudgments due to short-term abnormal fluctuations, thereby improving the accuracy and stability of predictions and guaranteeing the safe operation of wind turbines under complex weather conditions.

[0050] Furthermore, by statistically analyzing the time difference between each warning time point and the initial time in the risk prediction sequence, and using its standard deviation (time distribution degree) to measure the distribution characteristics of risk events within the prediction window: when the time distribution degree is greater than the preset distribution degree threshold, it indicates that the prediction model continuously provides high-risk signals throughout the entire prediction period, rather than sporadic or occasional isolated fluctuations; this cross-time period distribution width reflects the persistence and universality of fault evolution, thereby ensuring that the system only triggers warnings when it truly faces continuous risks, which avoids false alarms caused by single sudden predictions and can provide timely warnings when the overall risk spreads, significantly improving the reliability and timeliness of warnings.

[0051] Furthermore, by quantitatively analyzing the duration of continuous excessive vibration periods, the standard deviation (abnormal distribution degree) of the "distribution duration" can be extracted. This allows for the accurate capture of the repetitive-recovery-repetitive characteristics of blade vibration under gust impact. When the abnormal distribution degree exceeds a preset threshold, it indicates that the vibration anomaly is not merely a single short-term fluctuation, but rather a clustered response occurring multiple times over a long period. This precisely corresponds to the cumulative structural impact of strong winds and gusts on the unit. It can filter out normal small-amplitude flutter and issue timely warnings when repeated impacts truly threaten the stability of the unit, thereby significantly improving the accuracy and lead time of fault identification. Attached Figure Description

[0052] Figure 1 This is a flowchart of the extreme weather power generation equipment failure prediction method based on a time-series large model in this embodiment;

[0053] Figure 2 This is a logic diagram for determining abnormal lubrication trends in this embodiment;

[0054] Figure 3 This embodiment defines the logic diagram for determining the generator of interest;

[0055] Figure 4 The logic diagram for adjusting the preset gust impact threshold in this embodiment is shown. Detailed Implementation

[0056] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0057] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0058] Please see Figure 1The diagram shown is a flowchart of the extreme weather power generation equipment failure prediction method based on a time-series large model in this embodiment. This embodiment provides an extreme weather power generation equipment failure prediction method based on a time-series large model, including:

[0059] Real-time synchronous acquisition of the triaxial vibration peak value of the blades in the wind turbine of each wind turbine unit during strong winds, the oil viscosity of the gearbox input shaft, the current harmonic content of the generator stator, the starting torque increment of the yaw system, the wind speed change rate at the top of the nacelle, and the temperature rise rate of the main shaft bearing.

[0060] Based on the triaxial vibration peak value and the preset gust impact threshold within the preset anomaly determination period, an anomaly determination result is obtained;

[0061] Based on the anomaly determination result, the type of the abnormal event is determined to be a lubrication anomaly trend according to the oil viscosity change rate and the temperature rise rate, and the type determination result is obtained.

[0062] Based on the type determination result, according to the current harmonic content, the triaxial vibration peak value and the preset risk evolution threshold, the risk level of the fault trend is determined to be a severe warning, and the risk determination result is obtained.

[0063] Based on the risk assessment result, the preset gust impact threshold is adjusted according to the starting torque increment and the wind speed change rate in the next preset first adjustment cycle, and the risk evolution threshold is adjusted according to the number of times the gust impact threshold is adjusted in the next preset second adjustment cycle and the oil viscosity.

[0064] Based on all the risk determination results obtained after adjusting the risk evolution threshold within the preset prediction period, the triaxial vibration peak value, and the preset time series large model, a risk prediction sequence is generated, and an early warning alarm is issued based on the timestamp of the risk prediction sequence.

[0065] This embodiment employs an industrial Ethernet data acquisition system based on IEEE 1588 Precision Clock Synchronization (PTP). Various sensors are connected to an edge computing gateway within the nacelle via high-bandwidth fiber optic cables or shielded twisted-pair cables. Specifically, a MEMS triaxial accelerometer with a sampling frequency of 1kHz is installed in the middle of the blades; a microfluidic online viscometer is integrated into the gearbox input shaft oil circuit to output real-time oil viscosity values; a power quality analyzer is connected to the generator stator side to monitor current harmonic content; a torque sensor is installed on the yaw system bearing to record the increment of starting torque; an ultrasonic weather station or laser Doppler anemometer is deployed on the top of the nacelle to measure the rate of wind speed change; and a high-precision K-type thermocouple is installed in the bearing housing at the front end of the main shaft to detect the temperature rise rate. All sensor data are aligned to the nanosecond level using PTP-calibrated timestamps, preprocessed, and then uploaded by the edge gateway via fiber optic link to the central time-series large model server, achieving real-time synchronous acquisition and high-precision correlation analysis of parameters.

[0066] The preset anomaly judgment period is a sampling time window used to determine whether the vibration peak value exceeds the limit. It depends on the vibration response characteristics of the unit and the data sampling frequency, and is usually set between 10 seconds and 60 seconds. In this embodiment, it is set to 30 seconds, which can capture the vibration change caused by gust impact in time.

[0067] The preset risk evolution threshold is a comprehensive indicator threshold for risk level classification. It depends on the statistical distribution of multi-source characteristics (harmonic content, vibration amplitude, etc.) and is usually set between 0.7 and 0.9. In this embodiment, it is set to 0.85, which can effectively identify high-risk trends.

[0068] The preset first adjustment cycle is the time interval for adaptively adjusting the gust threshold, which depends on the dynamic rate of change of yaw torque and wind speed. It is usually set between 1 minute and 5 minutes. In this embodiment, it is set to 2 minutes to respond to sudden changes in wind conditions in real time.

[0069] The preset second adjustment cycle is the trigger cycle for correcting the risk evolution threshold. It depends on the threshold adjustment frequency and the oil viscosity fluctuation characteristics. It is usually set between 10 minutes and 30 minutes. In this embodiment, it is set to 15 minutes, which can balance the stability and sensitivity of the threshold.

[0070] The preset prediction period is the time range for generating risk prediction sequences. It depends on the model prediction accuracy and operation and maintenance decision requirements, and is usually set between 24 hours and 72 hours. In this embodiment, it is set to 48 hours, which can provide sufficient lead time for scheduling and deployment.

[0071] The preset time series large model is a deep time series network specifically designed for multi-step risk prediction based on multi-source monitoring data of wind turbine units. This embodiment employs a multi-channel self-attention prediction model based on the Informer architecture. Its structure and process include: taking all risk assessment results within a preset prediction period, adjusted for risk evolution thresholds, and the corresponding triaxial vibration peak value at the center of the blade at the corresponding moment, and interleaving them at the minute level to form an M×2 feature matrix, where M is the number of time steps (i.e., the total number of time points within the prediction period); adding channel identifier embeddings to the two columns of the feature matrix, and then superimposing sine-cosine position encoding to obtain an input tensor of shape M×D, where D is the hidden layer dimension (the length of the vector mapped to each time step); feeding the input tensor into a multi-layer self-attention encoder, which efficiently extracts long-short-term dependency features across time steps and channels through a sparse global attention mechanism; the decoder receives the feature representation output by the encoder and the "historical target" sequence (in this embodiment, the risk assessment results of the most recent 60 time steps), and uses a sliding window and multi-step recursive strategy to accurately capture the risk evolution inertia and trend; the decoder sequentially generates a risk probability value sequence for the next T time steps (in this embodiment, the preset prediction period T = 48 × 60 minutes), forming the final risk prediction sequence.

[0072] By introducing multiple parameters closely related to structural loads, electrical disturbances, and thermal-lubrication conditions, such as the triaxial vibration peak value in the middle of the blade, the oil viscosity of the gearbox, the current harmonics of the stator, the starting torque increment of the yaw system, the wind speed change rate at the top of the nacelle, and the temperature rise rate of the main shaft bearing, the system achieves full-chain monitoring of the wind turbine's operating status under strong wind extreme weather conditions. Gust impacts directly cause increased blade vibration, which is transmitted to the gearbox, causing fluctuations in oil viscosity and changes in the temperature rise rate. Lubrication degradation further leads to increased component friction, reflected in changes in yaw torque and stator current harmonics. The risk level assessment integrates the coupling trend of electrical and mechanical signals, thereby driving dynamic threshold adjustment and achieving adaptive optimization of the model. Finally, a continuous fault risk prediction sequence is generated through a time-series large model, which can accurately judge trends and provide early warnings under multi-source coupled disturbance conditions, improving the fault perception sensitivity and operational safety of wind power equipment. This effectively solves the problem of delayed identification of abnormal lubrication trends under extreme weather conditions due to incomplete monitoring parameter coverage and fixed threshold settings, resulting in untimely fault warnings and increased unit downtime losses.

[0073] Please see Figure 2 As shown, this is the logic diagram for determining the abnormal lubrication trend in this embodiment. In this embodiment, the process of determining the type of the abnormal event as an abnormal lubrication trend based on the oil viscosity change rate and the temperature rise rate, and obtaining the type determination result, includes:

[0074] The maximum-minimum normalization process is applied to all viscosity change rates to obtain a set of normalized viscosity change values, and the maximum-minimum normalization process is applied to all temperature rise rates to obtain a set of normalized temperature rise rate values.

[0075] Calculate the Pearson correlation coefficient between the normalized set of viscosity changes and the normalized set of temperature rise rates to obtain the magnitude of the correlation.

[0076] When the magnitude of the change is greater than a preset magnitude threshold, the type of the abnormal event is determined to be a lubrication abnormal trend, and the type determination result is obtained.

[0077] The preset correlation amplitude threshold refers to the critical value of the Pearson correlation coefficient used to determine the synchronicity between changes in oil viscosity and the rate of temperature rise. It depends on the correlation distribution of these two parameters under historical extreme weather conditions and is usually set between 0.6 and 0.8. In this embodiment, it is set to 0.7, which can effectively distinguish between the synchronous temperature rise trend caused by actual lubrication degradation and occasional unrelated fluctuations.

[0078] By normalizing the rate of change in oil viscosity and the rate of temperature rise in the spindle bearing (existing technology, not elaborated further), the two types of data are analyzed on a unified scale, thereby improving the accuracy of measuring the consistency of their trend changes. Using the Pearson correlation coefficient as a measure of the strength of the correlation helps to accurately determine whether the deterioration of lubrication performance and the temperature rise have a synchronous trend. When the two are highly positively correlated, it indicates that the decline in lubrication performance has significantly affected heat conduction and mechanical friction, exhibiting the typical characteristic of lubrication degradation leading to increased temperature rise. This method effectively reveals the linkage evolution law between lubrication state and thermal effect during the formation of abnormal lubrication trends, thereby enhancing the accuracy and explanatory power of type determination.

[0079] Specifically, the process of determining the risk level of the fault trend as a severe warning based on the current harmonic content, the triaxial vibration peak value, and the preset risk evolution threshold of each wind turbine includes:

[0080] Calculate the standard deviation of all current harmonic content from the initial time to each time within the preset level determined period, obtain several harmonic content fluctuation values, and form a set of harmonic content fluctuations.

[0081] Based on the set of harmonic content fluctuations between any two adjacent wind turbine generators, a number of generators of interest are identified from the wind turbine generator set.

[0082] Based on the triaxial vibration peak values ​​of any two adjacent generators of interest and a preset risk evolution threshold, the risk level of the fault trend is determined to be a severe warning.

[0083] The preset level determination period is a time window used to collect current harmonics and vibration fluctuation data to assess the risk level. It depends on the electrical and mechanical signal response rate of the wind turbine and is usually set between 1 minute and 10 minutes. In this embodiment, it is set to 5 minutes to balance the timeliness of early warning and the stability of data.

[0084] First, by calculating the standard deviation of the current harmonic content of each wind turbine within the judgment period, the degree of dynamic load instability caused by wind shear or control fluctuations on the electrical side is quantified. Then, the electrical fluctuation sets of adjacent wind turbines are compared to screen out "units of concern" with abnormal harmonic synchronization, so as to identify equipment that may be affected by the same local extreme wind field. Finally, by combining the triaxial vibration peak value of these units with the preset risk evolution threshold, the units most prone to failure under electromechanical dual stress can be accurately identified, avoiding missed judgments caused by simply relying on the overall average index, and issuing the most targeted severe warning in advance.

[0085] Please see Figure 3 As shown, this is the logic diagram for determining the generators of interest in this embodiment. The process of determining several generators of interest from the wind turbine generator set based on the harmonic content fluctuation set of any two adjacent wind turbine generators includes:

[0086] The Pearson correlation coefficient of the harmonic content fluctuation set of any two adjacent wind turbines is calculated to obtain the fluctuation consistency.

[0087] When the fluctuation consistency is greater than the preset consistency threshold, the corresponding wind turbine is determined to be a generator of interest, so as to identify a number of generators of interest from the wind turbine group.

[0088] The preset consistency threshold refers to the critical value of the Pearson correlation coefficient of the set of harmonic content fluctuations of adjacent wind turbines. It depends on the statistical distribution of the electrical response synchronicity of adjacent units under historical extreme weather conditions. It is usually set between 0.7 and 0.9. In this embodiment, it is set to 0.8, which can effectively distinguish between synchronous fluctuations caused by the combined effect of gusts and occasional fluctuations caused by random noise.

[0089] By calculating the correlation coefficient of harmonic fluctuations in the current of adjacent wind turbines, units exhibiting synchronous electrical responses under the same wind field disturbances can be identified. If neighboring units simultaneously experience harmonic amplitude fluctuations, it indicates they may be affected by similar gusts or load switching, making these units more prone to cascading effects on their structures and lubrication systems. By designating units with fluctuation consistency exceeding a threshold as "generators of concern," subsequent peak vibration analysis and risk assessment can focus on key areas truly impacted by extreme weather. This reduces the overall computational load and avoids misjudging local disturbances as overall field risks, significantly improving the accuracy and efficiency of fault warnings and responses.

[0090] Specifically, the process of determining the risk level of the fault trend as a severe warning based on the triaxial vibration peak values ​​of any two adjacent generators of interest and a preset risk evolution threshold includes:

[0091] Calculate the difference in the peak values ​​of the triaxial vibration of any two adjacent generators of interest to obtain several peak value variations;

[0092] Calculate the standard deviation of all peak value changes to obtain the peak value fluctuation.

[0093] When the peak change fluctuation value exceeds the preset risk evolution threshold, the risk level of the fault trend is determined to be a severe warning.

[0094] By focusing on the differences in the peak values ​​of triaxial vibration between generators, fluctuation characteristics representing the severity of structural response are extracted and their overall instability is quantified in the form of standard deviation. This is then compared with a preset risk evolution threshold. When the fluctuation exceeds the threshold, a severe warning level is triggered, reflecting the dynamic coordination relationship of the mechanical state between units. When the degree of disturbance of adjacent units under extreme weather conditions deviates significantly, it can be regarded as a direct signal of increased possibility of fault evolution, thereby realizing the early identification of fault trends and the judgment of risk escalation.

[0095] Please see Figure 4 As shown, this is a logic diagram for adjusting the preset gust impact threshold in this embodiment. In this embodiment, the process of adjusting the preset gust impact threshold according to the increment of the starting torque and the rate of change of wind speed within the next preset first adjustment cycle includes:

[0096] Calculate the standard deviation of all starting torque increments from the initial time to each time within the next preset first adjustment period to obtain several torque increment fluctuation values, and calculate the total wind speed change rate from the initial time to each time within the next preset first adjustment period to obtain several wind speed change fluctuation values.

[0097] Plot the curve of torque increment fluctuation over time to obtain the first curve, and plot the curve of wind speed change rate over time to obtain the second curve;

[0098] Vectorize the first curve to obtain the first vector, and vectorize the second curve to obtain the second vector;

[0099] Calculate the cosine similarity between the first and second vectors to obtain the adjustment synchronization degree;

[0100] When the adjustment synchronization degree is greater than the preset adjustment synchronization threshold, the preset gust impact threshold is increased according to the relative deviation between the adjustment synchronization degree and the preset adjustment synchronization threshold and the preset threshold adjustment coefficient, F'=F×[1+k×(S-S0) / S0], where F' is the increased preset gust impact threshold, F is the original preset gust impact threshold, S is the adjustment synchronization degree, S0 is the preset adjustment synchronization threshold, and k is the preset threshold adjustment coefficient.

[0101] The preset adjustment synchronization threshold is a benchmark value for measuring whether the fluctuation trend of wind speed change rate and starting torque increment is consistent in the time dimension. It depends on the cosine similarity level of the two in the normal response state in historical operating data. It is usually set between 0.6 and 0.85. In this embodiment, it is set to 0.75, which can effectively identify the consistency of system response and determine whether dynamic adjustment is needed.

[0102] The preset threshold adjustment coefficient is a proportional coefficient used to control the increase of the preset gust impact threshold when the wind speed change rate and the start-up torque increment are highly synchronous. It depends on the structural tolerance and response strategy requirements of the wind turbine under different meteorological change scenarios. It is usually set between 0.1 and 0.3. In this embodiment, it is set to 0.2, which can achieve fine control of the threshold and avoid over-adjustment that may cause misjudgment or missed alarm.

[0103] By introducing synchronization degree and cosine similarity as indicators to measure the dynamic coupling relationship between the rate of change of wind speed and the fluctuation of the increment of starting torque, this embodiment can effectively identify the consistency of the yaw system's response to wind speed disturbances. Furthermore, based on the relative deviation between the two, a set threshold adjustment coefficient is used to quantify the gain of the preset gust impact threshold. This not only avoids misjudging short-term synchronization fluctuations as anomalies but also achieves dynamic adjustment of the threshold, enabling the system to have adaptive adjustment capabilities. This method integrates temporal volatility and trend consistency at the numerical level, improving the ability to reasonably judge the operating boundaries of wind turbines under complex weather conditions, thereby enhancing the robustness of anomaly identification and the accuracy of the early warning threshold.

[0104] Specifically, the process of adjusting the risk evolution threshold based on the number of times the gust impact threshold is adjusted within the next preset second adjustment cycle and the oil viscosity includes:

[0105] Calculate the standard deviation of all the times the gust impact threshold is adjusted to obtain the fluctuation value of the number of adjustments, and calculate the standard deviation of all oil viscosity to obtain the fluctuation value of oil viscosity.

[0106] The risk evolution threshold is adjusted based on the fluctuation values ​​of the number of adjustments and the fluctuation values ​​of oil viscosity.

[0107] By introducing a strategy of co-correcting the risk evolution threshold using the fluctuation values ​​of adjustment frequency and oil viscosity, the operational stability and lubrication status changes of wind turbine units during strong winds can be dynamically reflected. Frequent adjustments to the gust impact threshold, coupled with increased volatility, often indicate severe external wind field disturbances or abnormal control system response. Simultaneously, oil viscosity fluctuations reflect the stress and thermal state changes in the gearbox lubrication system. Using the standard deviation of both as a quantitative basis to adjust the risk evolution threshold not only enhances the system's sensitivity to potential fault risks but also effectively filters out occasional disturbances, improving the stability and robustness of the early warning mechanism. Furthermore, by establishing a linkage mechanism between the control system's adjustment frequency and key mechanical performance indicators, adaptive optimization of the fault risk trend judgment threshold can be achieved.

[0108] Specifically, the process of adjusting the risk evolution threshold based on the fluctuation value of the number of adjustments and the fluctuation value of the oil viscosity includes:

[0109] Calculate the absolute difference between the fluctuation value of the number of adjustments and the preset fluctuation threshold to obtain the fluctuation deviation of the number of adjustments; and calculate the absolute difference between the fluctuation value of the oil viscosity and the preset viscosity fluctuation threshold to obtain the viscosity fluctuation deviation.

[0110] When the frequency fluctuation deviation is greater than the preset frequency fluctuation deviation threshold and the viscosity fluctuation deviation is greater than the preset viscosity fluctuation deviation threshold, the risk evolution threshold is increased according to the relative deviation between the frequency fluctuation deviation and the preset frequency fluctuation deviation threshold, and the relative deviation between the viscosity fluctuation deviation and the preset viscosity fluctuation deviation threshold. Y'=Y×{1+u×[i×(H-H0) / H0+j×(P-P0) / P0]}, where Y' is the increased risk evolution threshold, Y is the original risk evolution threshold, u is the preset risk adjustment coefficient, i is the preset frequency deviation weight, H is the frequency fluctuation deviation, H0 is the preset frequency fluctuation deviation threshold, j is the preset viscosity deviation weight, P is the viscosity fluctuation deviation, and P0 is the preset viscosity fluctuation deviation threshold.

[0111] The preset fluctuation threshold is a benchmark value used to determine whether the fluctuation of the number of times the wind turbine adjusts the impact threshold is abnormal. It depends on the historical statistical characteristics of the wind turbine's adjustment frequency under extreme weather conditions and is usually set between 1 and 5 times. In this embodiment, it is set to 3 times, which can effectively distinguish between normal fluctuations and abnormal adjustment behavior.

[0112] The preset viscosity fluctuation threshold is a threshold used to determine whether the viscosity fluctuation of the oil exceeds the normal range. It depends on the performance changes of the lubricating oil under different operating conditions and is usually set between 0.01 and 0.05. In this embodiment, it is set to 0.03, which can accurately reflect the abnormal trend of the lubrication system.

[0113] The preset fluctuation deviation threshold is the maximum allowable deviation between the fluctuation value of the number of adjustments and the preset fluctuation threshold. It depends on the fluctuation amplitude of the actual operation of the equipment and is usually set between 0.5 and 1.5. In this embodiment, it is set to 1.0, which can reasonably define the difference between normal and abnormal fluctuations.

[0114] The preset viscosity fluctuation deviation threshold is the maximum tolerable deviation between the oil viscosity fluctuation value and the preset viscosity fluctuation threshold. It depends on the stability of the lubrication system fluctuation and is usually set between 0.005 and 0.02. In this embodiment, it is set to 0.01 to ensure the scientific and accurate adjustment of the threshold.

[0115] The preset risk adjustment coefficient is a proportional factor used to adjust the increase of the risk evolution threshold. It depends on the risk sensitivity requirements of wind power equipment and is usually set between 0.1 and 0.5. In this embodiment, it is set to 0.3 to balance the timeliness and stability of risk response.

[0116] The preset number of deviation weight is used to weight the influence of the number of adjustment fluctuations on the risk threshold adjustment. It depends on the contribution of the number of fluctuations to the fault risk and is usually set between 0.4 and 0.7. In this embodiment, it is set to 0.6 to strengthen the role of the number of adjustments in risk judgment.

[0117] The preset viscosity deviation weight is a weighting factor used to measure the proportion of oil viscosity fluctuation deviation in the overall risk evolution threshold adjustment, and depends on the degree of influence of the lubrication system on the operational stability of the power generation equipment. It is usually set between 0.3 and 0.6, and in this embodiment it is set to 0.4, which can effectively reflect the important role of abnormal lubrication performance in the overall risk assessment, thereby improving the response accuracy of fault prediction and the rationality of regulation.

[0118] By comprehensively analyzing the relative changes in the frequency of adjustment fluctuations and the viscosity fluctuations of the lubrication system, the risk evolution threshold is appropriately increased to dynamically adapt to the operating status of wind turbines under extreme weather conditions. The frequency of adjustment reflects the frequency of gust impact threshold adjustments, while viscosity fluctuations reflect the stability of the lubrication system; both jointly affect the risk level of equipment failure development. By introducing a preset risk adjustment coefficient and corresponding weights, the two deviations are weighted and calculated to achieve precise adjustment of the risk threshold. This ensures that the fault early warning system can sensitively capture potential risks while avoiding misjudgments due to short-term abnormal fluctuations, thereby improving the accuracy and stability of predictions and guaranteeing the safe operation of wind turbines under complex weather conditions.

[0119] Specifically, the process of issuing early warning alerts based on the timestamps of the risk prediction sequence includes:

[0120] Calculate the difference between each of the corresponding timestamps and the initial time in the risk prediction sequence to obtain several distribution durations, and calculate the standard deviation of all distribution durations to obtain the time distribution degree;

[0121] When the time distribution degree is greater than the preset distribution degree threshold, it is determined that the wind turbine unit has malfunctioned and the aforementioned early warning alarm is issued.

[0122] The preset distribution threshold is the standard deviation critical value used to determine the dispersion of risk prediction time points. It depends on the time distribution fluctuation characteristics of historical fault events within the prediction window and is usually set between 300 seconds and 900 seconds. In this embodiment, it is set to 600 seconds, which can effectively distinguish between occasional short-term high-risk signals and continuous fault evolution trends.

[0123] By statistically analyzing the time difference between each warning time point and the initial time in the risk prediction sequence, and using its standard deviation (time distribution degree) to measure the distribution characteristics of risk events within the prediction window: when the time distribution degree is greater than the preset distribution degree threshold, it indicates that the prediction model continuously provides high-risk signals throughout the entire prediction period, rather than sporadic or occasional isolated fluctuations; this cross-time period distribution width reflects the persistence and universality of fault evolution, thereby ensuring that the system only triggers warnings when it truly faces continuous risks, which avoids false alarms caused by single sudden predictions and can provide timely warnings when the overall risk spreads, significantly improving the reliability and timeliness of warnings.

[0124] Specifically, the process of determining the existence of an abnormal event based on the triaxial vibration peak value and the preset gust impact threshold within a preset anomaly determination period, and obtaining the anomaly determination result, includes:

[0125] When the peak value of triaxial vibration within the preset anomaly judgment period is greater than the preset vibration peak value threshold, the current timestamp is recorded, and recording stops when the peak value of triaxial vibration is less than or equal to the preset vibration peak value threshold, thus obtaining several durations.

[0126] Obtain the time period from the initial moment of the preset anomaly judgment period to the timestamp at the center of each duration to obtain several distributed durations;

[0127] Calculate the standard deviation of all the distribution durations to obtain the anomaly distribution degree;

[0128] When the abnormal distribution degree is greater than the preset gust impact threshold, the existence of the abnormal event is determined, and the abnormal determination result is obtained.

[0129] The preset vibration peak threshold is the peak critical value for abnormal vibration determined by the triaxial accelerometer in the middle of the blade. It depends on the historical gust peak statistics and the blade structure tolerance, and is usually set between 1.5g and 4g. In this embodiment, it is set to 3g, which can effectively distinguish between normal operating flutter and abnormal large-amplitude vibration caused by gust impact.

[0130] The preset gust impact threshold refers to the critical value of the standard deviation of the duration of abnormal vibration caused by continuous gusts within the judgment period. It depends on the statistical distribution of the duration of abnormal blade vibration during historical typhoons and is usually set between 5 and 20 seconds. In this embodiment, it is set to 10 seconds, which can distinguish between normal short-term vibration fluctuations and long-term vibration distribution deviations caused by gust impacts.

[0131] By quantitatively analyzing the duration of continuous excessive vibration periods and extracting the standard deviation (abnormal distribution degree) of the "distribution duration," the repetitive-recovery-repetitive characteristics of blade vibration under gust impact can be accurately captured. When the abnormal distribution degree exceeds a preset threshold, it indicates that the vibration anomaly is not just a single short-term fluctuation, but a clustered response that occurs multiple times over a long period of time. This corresponds precisely to the cumulative structural impact of strong winds and gusts on the unit. It can filter out normal small-amplitude flutter and issue timely warnings when repeated impacts truly threaten the stability of the unit, thereby significantly improving the accuracy and lead time of fault identification.

[0132] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for predicting power generation equipment failures in extreme weather based on a time-series large model, characterized in that, include: Real-time synchronous acquisition of the triaxial vibration peak value of the blades in the wind turbine of each wind turbine unit during strong winds, the oil viscosity of the gearbox input shaft, the current harmonic content of the generator stator, the starting torque increment of the yaw system, the wind speed change rate at the top of the nacelle, and the temperature rise rate of the main shaft bearing. Based on the triaxial vibration peak value and the preset gust impact threshold within the preset anomaly determination period, an anomaly determination result is obtained; Based on the anomaly determination result, the type of the abnormal event is determined to be a lubrication anomaly trend according to the oil viscosity change rate and the temperature rise rate, and the type determination result is obtained. Based on the type determination result, according to the current harmonic content, the triaxial vibration peak value and the preset risk evolution threshold, the risk level of the fault trend is determined to be a severe warning, and the risk determination result is obtained. Based on the risk assessment result, the preset gust impact threshold is adjusted according to the starting torque increment and the wind speed change rate in the next preset first adjustment cycle, and the risk evolution threshold is adjusted according to the number of times the gust impact threshold is adjusted in the next preset second adjustment cycle and the oil viscosity. Based on all the risk determination results obtained after adjusting the risk evolution threshold within the preset prediction period, the triaxial vibration peak value, and the preset time series large model, a risk prediction sequence is generated, and an early warning alarm is issued based on the timestamp of the risk prediction sequence.

2. The method for predicting power generation equipment failures in extreme weather based on a large time-series model according to claim 1, characterized in that, The process of determining the type of the abnormal event as a lubrication abnormality trend based on the oil viscosity change rate and the temperature rise rate, and obtaining the type determination result, includes: The change rate of the oil viscosity and the rate of temperature rise within the period are determined based on the preset abnormal trend to determine the relevant magnitude of the change; Based on the comparison between the magnitude of the change and the preset magnitude threshold, the type of the abnormal event is determined to be a lubrication abnormal trend, and the type determination result is obtained.

3. The method for predicting power generation equipment failures in extreme weather based on a large time-series model according to claim 2, characterized in that, The process of determining the risk level of the fault trend as a severe warning based on the current harmonic content, the triaxial vibration peak value, and the preset risk evolution threshold of each wind turbine includes: The harmonic content fluctuation set is determined based on the current harmonic content from the initial time to each time within the preset level. Based on the set of harmonic content fluctuations between any two adjacent wind turbine generators, a number of generators of interest are identified from the wind turbine generator set. Based on the triaxial vibration peak values ​​of any two adjacent generators of interest and a preset risk evolution threshold, the risk level of the fault trend is determined to be a severe warning.

4. The method for predicting power generation equipment failures in extreme weather based on a large time-series model according to claim 3, characterized in that, The process of identifying several generators of interest from the wind turbine array based on the set of harmonic content fluctuations between any two adjacent wind turbines includes: The consistency of fluctuations is determined based on the two sets of harmonic content fluctuations. Based on the comparison between the fluctuation consistency and the preset consistency threshold, several generators of interest are identified from the wind turbine units.

5. The method for predicting power generation equipment failures in extreme weather based on a large time-series model according to claim 4, characterized in that, The process of determining the risk level of the fault trend as a severe warning based on the triaxial vibration peak values ​​of any two adjacent generators of interest and a preset risk evolution threshold includes: Several peak value variations are determined based on the triaxial vibration peak values ​​of any two adjacent generators of interest; The peak change fluctuation value is determined based on all the aforementioned peak change values; Based on the comparison between the peak change fluctuation value and the preset risk evolution threshold, the risk level of the fault trend is determined to be a severe warning.

6. The method for predicting power generation equipment failures in extreme weather based on a large time-series model according to claim 5, characterized in that, The process of adjusting the preset gust impact threshold based on the increment of the starting torque and the rate of change of wind speed within the next preset first adjustment cycle includes: The synchronization degree is determined based on all the aforementioned starting torque increments and all the aforementioned wind speed change rates; The preset gust impact threshold is adjusted based on the comparison between the adjusted synchronization degree and the preset adjusted synchronization threshold.

7. The method for predicting power generation equipment failures in extreme weather based on a large time-series model according to claim 6, characterized in that, The process of adjusting the risk evolution threshold based on the number of times the gust impact threshold is adjusted within the next preset second adjustment cycle and the oil viscosity includes: The fluctuation value of the number of adjustments is determined based on the total number of times the gust impact threshold is adjusted, and the fluctuation value of the oil viscosity is determined based on the total oil viscosity. The risk evolution threshold is adjusted based on the fluctuation value of the number of adjustments and the fluctuation value of the oil viscosity.

8. The method for predicting power generation equipment failures in extreme weather based on a large time-series model according to claim 7, characterized in that, The process of adjusting the risk evolution threshold based on the fluctuation value of the adjustment frequency and the fluctuation value of the oil viscosity includes: The fluctuation deviation is determined based on the fluctuation value of the adjustment number and the preset fluctuation threshold, and the fluctuation is also determined based on the oil viscosity fluctuation. The viscosity fluctuation deviation is determined by the value and the preset viscosity fluctuation threshold; The risk evolution threshold is adjusted based on the comparison results of the frequency fluctuation deviation and the preset frequency fluctuation deviation threshold, and the comparison results of the viscosity fluctuation deviation and the preset viscosity fluctuation deviation threshold.

9. The method for predicting power generation equipment failures in extreme weather based on a large time-series model according to claim 8, characterized in that, The process of issuing early warning alerts based on the timestamps of risk prediction sequences includes: Determine the degree of time distribution based on all timestamps; Based on the comparison between the time distribution degree and the preset distribution degree threshold, it is determined that the wind turbine unit has malfunctioned, and the early warning alarm is issued.

10. The method for predicting power generation equipment failures in extreme weather based on a large time-series model according to claim 9, characterized in that, The process of determining the existence of an abnormal event based on the triaxial vibration peak value and the preset gust impact threshold within the preset anomaly determination period, and obtaining the anomaly determination result, includes: The degree of abnormal distribution is determined based on the comparison between the triaxial vibration peak value and the preset vibration peak value threshold. The abnormal event is determined to exist based on the comparison between the abnormal distribution degree and the preset gust impact threshold, and the abnormality determination result is obtained.

Images