Root cause positioning method and system for hydraulic variable pitch angle execution deviation

Abstract

The embodiment of the invention provides a root cause positioning method and system for hydraulic variable pitch angle execution deviation. The method comprises the steps that operation data of a wind generating set are collected in real time; target feature extraction is executed based on the operation data, and the target feature extraction comprises complex wind regime feature extraction, hydraulic driving system feature extraction and blade coupling vibration feature extraction; when a variable pitch tracking fault is triggered, root cause classification is executed based on the extraction result of the complex wind regime feature extraction, the extraction result of the hydraulic driving system feature extraction and the extraction result of the blade coupling vibration feature extraction; the root cause classification comprises complex wind conditions, hydraulic actuating mechanism abnormity, blade coupling vibration and unclassified abnormity. According to the method, three types of root causes of variable-pitch driving overrun caused by complex wind conditions, leakage / pressure abnormity of the hydraulic executing mechanism and angle oscillation caused by coupling vibration of the blades are accurately distinguished, and the diagnosis precision is effectively improved.

Description

Technical Field

[0001] The present application belongs to the field of wind power generation technology, and specifically relates to a method and system for locating the root cause of hydraulic pitch angle execution deviation. Background Art

[0002] Wind turbine pitch tracking problems can have multiple detrimental effects on wind turbine performance, component life, and grid stability. The pitch system needs to adjust the blade angle in real time according to wind speed and power generation to achieve optimal wind energy capture. Inaccurate tracking will lead to reduced wind energy utilization efficiency and reduced power generation. At the same time, the blades are subjected to uneven aerodynamic loads, which exacerbates local wear and shortens their service life. In severe cases, they may deform, crack, or even break, causing safety accidents and economic losses. In addition, pitch angle deviations can also cause fluctuations in unit power output. When this problem occurs in multiple units at the same time, it may cause grid voltage fluctuations and frequency deviations, affecting the stable operation of the grid, and even triggering unit overspeed, overload and other protection mechanisms, resulting in emergency shutdowns.

[0003] Pitch systems primarily include hydraulic and electric drives. The causes of inaccurate pitch tracking are complex and diverse. For hydraulic pitch systems, hydraulic pump failure, pipeline leakage, or hydraulic cylinder seal damage can lead to unstable system pressure or insufficient drive force, affecting pitch accuracy. Defects in the pitch control algorithm (such as PID parameter mismatch) can cause adjustment overshoot, oscillation, or response delays. Furthermore, complex wind conditions (such as strong winds) can increase the aerodynamic drag of the blades, causing the pitch drive to exceed limits and prevent accurate tracking.

[0004] Currently, when pitch tracking deviations consistently exceed thresholds, turbines typically shut down for protection and rely on technicians to manually analyze massive amounts of operating data to locate the root cause. Two types of fault diagnosis methods exist in existing technologies: one converts fault data into feature vectors and determines the root cause based on unsupervised learning and matching similarities; the other determines system anomalies by calculating the residual error of the wind energy utilization coefficient and then analyzes the cause of the anomaly based on pitch angle / torque consistency. However, the former method fails to incorporate the system's operating mechanism, while the latter lacks a specific root cause classification mechanism, making it difficult to accurately and efficiently locate faults.

[0005] Therefore, there is an urgent need for a diagnostic method based on the operating mechanism of the hydraulic pitch system that can automatically identify three typical root causes (complex wind conditions, hydraulic actuator abnormalities, and blade coupling vibrations) to improve operation and maintenance efficiency. Summary of the Invention

[0006] The present application proposes a method and system for locating the root cause of hydraulic pitch angle execution deviation, which is used to solve the defects of the above-mentioned prior art.

[0007] According to a first aspect of an embodiment of the present application, a method for locating the root cause of a hydraulic pitch angle execution deviation is provided, comprising:

[0008] Real-time collection of wind turbine operating data;

[0009] Based on the operating data, target feature extraction is performed, wherein the target feature extraction includes complex wind condition feature extraction, hydraulic drive system feature extraction, and blade coupled vibration feature extraction;

[0010] When a pitch tracking fault is triggered, root cause classification is performed based on the extraction results of the complex wind condition feature extraction, the extraction results of the hydraulic drive system feature extraction, and the extraction results of the blade coupled vibration feature extraction. The root cause classification includes complex wind conditions, hydraulic actuator abnormalities, blade coupled vibrations, and unclassified abnormalities.

[0011] In some embodiments, the operating data includes: wind speed sampling information, pitch angle command values, actual execution values ​​of pitch angles of three blades, hydraulic station pressure values, and tank liquid level values. Executing target feature extraction based on the operating data includes:

[0012] Calculating the average wind speed and turbulence intensity within a predetermined time window based on the wind speed sampling information to perform complex wind condition feature extraction;

[0013] Calculating a pressure deviation between the real-time pressure of the hydraulic station and a preset pressure, and a level deviation between the real-time level of the oil tank and a preset minimum level, based on the hydraulic station pressure value and the oil tank liquid level value, so as to perform hydraulic drive system feature extraction;

[0014] According to the pitch angle command value and the actual execution values ​​of the pitch angles of the three blades, it is respectively detected whether the actual execution values ​​of the pitch angles of the three blades oscillate around the pitch angle command value, and the fluctuation frequency characteristics of the blades are extracted during oscillation to perform the blade coupled vibration feature extraction.

[0015] In some embodiments, calculating the average wind speed and turbulence intensity within a predetermined time window based on the wind speed sampling information to perform complex wind condition feature extraction includes:

[0016] Acquiring the wind speed sampling information within a continuous time window;

[0017] The average wind speed and standard deviation of the wind speed sampling information are calculated, and the standard deviation is used as the turbulence intensity.

[0018] In some embodiments, the turbulence intensity calculated within a predetermined time window may be expressed by the following formula:

[0019]

[0020] Wherein, Tur represents the turbulence intensity, Vi Indicates the real-time wind speed sampling sequence, i represents the sequence number, V m represents the average wind speed, and 3000 is the number of sampling points.

[0021] In some embodiments, when a pitch tracking fault is triggered, performing root cause classification based on the extraction results of the complex wind condition feature extraction, the extraction results of the hydraulic drive system feature extraction, and the extraction results of the blade coupled vibration feature extraction includes:

[0022] If the average wind speed exceeds a first threshold and the turbulence intensity exceeds a second threshold, determining that the root cause is the complex wind condition;

[0023] If the pressure deviation exceeds a third threshold or the liquid level deviation exceeds a fourth threshold, it is determined that the root cause is an abnormality of the hydraulic actuator;

[0024] If the deviation between the extracted fluctuation frequency characteristic and the blade natural frequency characteristic is within a preset tolerance range, determining that the root cause is the blade coupled vibration;

[0025] If none of the above conditions are met, it is determined that the root cause is the unclassified abnormality.

[0026] In some embodiments, before calculating the pressure deviation value between the real-time pressure of the hydraulic station and the preset design pressure, and the liquid level deviation value between the real-time liquid level of the oil tank and the preset minimum liquid level, the method includes:

[0027] The hydraulic station pressure value and the oil tank liquid level value are subjected to low-pass filtering to filter out high-frequency noise signals.

[0028] In some embodiments, the blade coupled vibration feature extraction includes:

[0029] When it is detected that the actual execution value of the pitch angle of any blade continuously oscillates within a preset range around the pitch angle command value, performing Fourier transform on the oscillation signal;

[0030] The transformed main frequency component is extracted as the fluctuation frequency feature.

[0031] In some embodiments, the root cause classification is performed on recorded data within a predetermined time period before the pitch tracking fault occurs, and the predetermined time period is 1 minute.

[0032] In some embodiments, the preset tolerance range of the blade coupled vibration is ±5% of the blade natural frequency.

[0033] According to a second aspect of an embodiment of the present application, a system for locating the root cause of a hydraulic pitch angle execution deviation is provided, comprising:

[0034] Data acquisition module, used to collect real-time operating data of wind turbines;

[0035] A feature extraction module, configured to perform target feature extraction based on the operating data, wherein the target feature extraction includes complex wind condition feature extraction, hydraulic drive system feature extraction, and blade coupled vibration feature extraction;

[0036] A root cause classification module is used to perform root cause classification based on the extraction results of the complex wind condition feature extraction, the extraction results of the hydraulic drive system feature extraction, and the extraction results of the blade coupled vibration feature extraction when a pitch tracking fault is triggered. The root cause classification includes complex wind conditions, hydraulic actuator abnormalities, blade coupled vibrations, and unclassified abnormalities.

[0037] According to a third aspect of an embodiment of the present application, a computer-readable storage medium is provided, on which a computer program is stored, characterized in that when the program is executed by a processor, the root cause locating method of the hydraulic pitch angle execution deviation is implemented.

[0038] The beneficial effects of the method and system for locating the root cause of hydraulic pitch angle execution deviation according to the embodiments of the present application include at least:

[0039] The embodiment of the present application simultaneously executes complex wind condition feature extraction, hydraulic drive feature extraction and blade vibration feature extraction to accurately distinguish the three root causes of pitch drive overrun caused by complex wind conditions, hydraulic actuator leakage / pressure abnormality and angle oscillation caused by blade coupled vibration, thereby effectively improving the diagnosis accuracy; by writing the feature extraction and root cause classification algorithm into the main control program, when the pitch tracking fault is triggered, the recorded data of the preset time before the fault is automatically analyzed, the root cause classification results are output in real time, and the manual analysis link is eliminated, which significantly shortens the single fault location time and effectively optimizes the operation and maintenance efficiency; the present application can also trigger targeted protection mechanisms based on the root cause classification results, improve active protection linkage, and enhance system reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a flow chart of a method for locating the root cause of a hydraulic pitch angle execution deviation according to an embodiment of the present application;

[0041] Figure 2 This is a structural schematic diagram of a system for locating the root cause of hydraulic pitch angle execution deviation according to an embodiment of the present application. DETAILED DESCRIPTION

[0042] In order to enable those skilled in the art to better understand the technical solution of the present application, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0043] The following detailed description of the embodiments of the present application is provided in conjunction with the accompanying drawings and examples. The following detailed description of the embodiments and the accompanying drawings are used to illustrate the principles of the present application, but are not intended to limit the scope of the present application, that is, the present application is not limited to the described embodiments.

[0044] Refer to the attached Figure 1 As shown, the embodiment of the present application discloses specific implementation steps of a method for locating the root cause of a hydraulic pitch angle execution deviation. The method is implemented by a system for locating the root cause of a hydraulic pitch angle execution deviation. The system includes a readable storage medium to ensure that those skilled in the art can implement the technical solution of the present application accordingly. The method specifically includes the following steps 110-130.

[0045] Step 110: collecting the operating data of the wind turbine generator set in real time.

[0046] Operational data refers to key operational data for the target wind farm units. This data can be collected through data monitoring by a programmable logic controller (PLC), with a sampling interval of 20ms. This operational data includes at least: wind speed sampling information, pitch angle command values, actual pitch angle values ​​for the three blades, hydraulic station pressure, and fuel tank level.

[0047] The wind speed sampling information can be a real-time wind speed sampling sequence (V i ), which can be the real-time continuous wind speed value measured by the anemometer installed on the top of the nacelle. The pitch angle command value (PAt) can be the target pitch angle generated by the main control system according to the wind speed and power optimization algorithm. The actual execution value of the pitch angle (PA1i, PA2i, PA3i) is the actual pitch position of the three blades fed back by the angle sensor. The hydraulic station pressure value (P0i) is the real-time working pressure of the hydraulic pitch system drive cylinder. The tank level value (Li) is the real-time height of the lubricating oil in the hydraulic tank.

[0048] Step 120 : performing target feature extraction based on the operating data. The target feature extraction includes complex wind condition feature extraction, hydraulic drive system feature extraction, and blade coupled vibration feature extraction.

[0049] In some embodiments, after collecting the operating data of the wind turbine generator set in real time, the process also includes reading the preset data of the wind turbine generator set, mainly including: reading the preset pressure (P0d) of the hydraulic station, the preset minimum liquid level (Ld) of the oil tank, and the blade natural frequencies (f1d, f2d, f3d, ..., fNd). The preset pressure of the hydraulic station is the minimum stable pressure value required for the hydraulic pitch system to maintain the rated output torque, the preset minimum liquid level of the oil tank is the critical liquid level height to ensure normal oil suction of the hydraulic pump and avoid cavitation, and the blade natural frequency is the frequency corresponding to the Nth order main vibration mode of the blade in the free state.

[0050] In some embodiments, after collecting the operating data of the wind turbine generator set in real time, the process also includes reading threshold setting values. The threshold setting values ​​may include, for example, an average wind speed threshold (FZVm), an extreme turbulence threshold (FZtur), and hydraulic station operating pressure deviation thresholds (FZp0, FZL). The average wind speed threshold is the minimum average wind speed critical value that triggers the determination of abnormal wind conditions, the extreme turbulence threshold is the upper limit of turbulence intensity for determining complex wind conditions, and the hydraulic station pressure deviation threshold is the maximum negative deviation allowed between the real-time pressure and the design value.

[0051] In some embodiments, based on the operating data, performing target feature extraction includes: calculating the average wind speed and turbulence intensity within a predetermined time window based on the wind speed sampling information to perform complex wind condition feature extraction; calculating the pressure deviation value between the real-time pressure of the hydraulic station and the preset pressure, and the level deviation value between the real-time tank level and the preset minimum level based on the hydraulic station pressure value and the oil tank level value to perform hydraulic drive system feature extraction; based on the pitch angle command value and the actual execution value of the pitch angle of the three blades, respectively detecting whether the actual execution value of the pitch angle of the three blades oscillates around the pitch angle command value, and extracting the fluctuation frequency characteristics of the blades during oscillation to perform blade coupled vibration feature extraction.

[0052] The complex wind condition feature extraction is primarily used to determine whether complex wind conditions are present. For example, if the wind speed fluctuates dramatically, exceeding the load design limit, the pitch drive device will be unable to meet the overload, resulting in hydraulic pitch execution deviation.

[0053] In an exemplary embodiment, based on the wind speed sampling information, the average wind speed and turbulence intensity within a predetermined time window are calculated to perform complex wind condition feature extraction, including: obtaining wind speed sampling information within a continuous time window; calculating the average wind speed and standard deviation of the wind speed sampling information, and using the standard deviation as the turbulence intensity.

[0054] The turbulence intensity within a predetermined time window can be calculated using the following formula:

[0055]

[0056] Among them, V m =average(V1+V2+…+V3000);

[0057] Where Tur represents the turbulence intensity, V i Indicates the real-time wind speed sampling sequence, i represents the sequence number, V m represents the average wind speed, 3000 is the number of sampling points, and average represents the averaging function. The above formula can be used to continuously extract sampling values ​​for 1 minute (for example, 3000 samples per minute).

[0058] In addition, we continue to compare and calculate whether the above turbulence intensity exceeds the turbulence limit threshold. This can be expressed by the following formula:

[0059] Turdet=Tur-FZtur;

[0060] Where Turdet is the difference between the turbulence intensity and the turbulence limit threshold.

[0061] In an exemplary embodiment, before calculating the pressure deviation value between the real-time pressure of the hydraulic station and the preset design pressure, and the level deviation value between the real-time liquid level of the oil tank and the preset minimum liquid level, it includes: performing low-pass filtering on the hydraulic station pressure value and the oil tank liquid level value to filter out high-frequency noise signals.

[0062] Among them, the judgment of the hydraulic drive system feature extraction is mainly used to judge whether the pressure of the hydraulic station is normal and whether the oil is within the normal range. If there are problems such as oil drop or pressure drop, it will not provide sufficient power for the pitch system, resulting in hydraulic pitch execution deviation.

[0063] For example, the hydraulic station pressure value P0i and the tank level value Li are filtered and passed through a low-pass filter to filter out high-frequency signals caused by changes in the switching value; the pressure deviation value (P0deti) between the real-time pressure of the hydraulic station and the preset pressure is calculated, and the level deviation value (Ldeti) between the real-time tank level and the preset minimum level is calculated. The pressure deviation value (P0deti) and the level deviation value (Ldeti) can be obtained, for example, by the following formula:

[0064] P0deti=P0i-P0d;

[0065] Ldeti=Li-Ld.

[0066] In an exemplary embodiment, blade coupled vibration feature extraction includes: when it is detected that the actual execution value of the pitch angle of any blade continuously oscillates within a preset range around the pitch angle command value, performing Fourier transform on the oscillation signal; and extracting the transformed main frequency component as the fluctuation frequency feature.

[0067] The extraction of blade coupled vibration features primarily determines whether coupled vibrations can cause the three blades of the variable pitch system to experience fluctuating torques, leading to unstable pitch angle tracking. This coupled vibration manifests itself as a spectrum that aligns with the blade's inherent characteristic frequency. Therefore, the presence of coupled vibrations is determined by checking whether the fluctuating frequency aligns with the blade's designed inherent frequency.

[0068] For example, we can first determine whether the actual execution value of the pitch angle of any blade fluctuates around the pitch command value; if the fluctuation is within the preset range, Fourier analysis is performed on the pitch angle signals PA1i, PA2i, and PA3i, and the 1-minute data is analyzed to extract the characteristic frequency of the fluctuation and obtain the transformed main frequency components f1, f2...fN.

[0069] In some embodiments, the preset tolerance range of the blade coupled vibration is ±5% of the blade natural frequency.

[0070] Step 130, when a pitch tracking fault is triggered, root cause classification is performed based on the extraction results of complex wind condition feature extraction, the extraction results of hydraulic drive system feature extraction, and the extraction results of blade coupled vibration feature extraction. The root cause classification includes complex wind conditions, hydraulic actuator abnormalities, blade coupled vibrations, and unclassified abnormalities.

[0071] In some embodiments, when a pitch tracking fault is triggered, root cause classification is performed based on the extraction results of complex wind condition feature extraction, the extraction results of hydraulic drive system feature extraction, and the extraction results of blade coupled vibration feature extraction, including: if the average wind speed exceeds a first threshold and the turbulence intensity exceeds a second threshold, it is determined that the root cause is a complex wind condition. For example, it can be determined that when Vm>FZVm and Turdet>FZtur, a pitch tracking fault caused by complex wind conditions is reported; if the pressure deviation exceeds a third threshold or the liquid level deviation exceeds a fourth threshold, it is determined that the root cause is an abnormality of the hydraulic actuator. For example, it is determined that P0deti>FZp0 or Ldeti>FZL occurs, then Report a pitch tracking fault caused by control abnormality or oil leakage in the hydraulic pitch system actuator; if the deviation between the extracted fluctuation frequency characteristics and the blade natural frequency characteristics is within the preset tolerance range, it is determined that the root cause is blade coupling vibration. For example, f1, f2···fN can be compared with the blade natural frequencies f1d, f2d···fNd. If the deviation between the extracted characteristic frequency and the natural frequency is within 5%, it is reported that there is a pitch tracking fault caused by blade coupling vibration; if none of the above conditions are met, it is determined that the root cause is an unclassified abnormality, such as reporting a pitch tracking fault caused by non-complex wind conditions, pitch actuator and blade coupling, and further investigation of the cause is required.

[0072] In some embodiments, the root cause classification is performed on the recorded data within a predetermined time period before the pitch tracking fault occurs, where the predetermined time period is 1 minute.

[0073] The embodiment of the present application simultaneously executes complex wind condition feature extraction, hydraulic drive feature extraction and blade vibration feature extraction to accurately distinguish the three root causes of pitch drive overrun caused by complex wind conditions, hydraulic actuator leakage / pressure abnormality and angle oscillation caused by blade coupled vibration, thereby effectively improving the diagnosis accuracy; by writing the feature extraction and root cause classification algorithm into the main control program, when the pitch tracking fault is triggered, the recorded data of the preset time before the fault is automatically analyzed, the root cause classification results are output in real time, and the manual analysis link is eliminated, which significantly shortens the single fault location time and effectively optimizes the operation and maintenance efficiency; the present application can also trigger targeted protection mechanisms based on the root cause classification results, improve active protection linkage, and enhance system reliability.

[0074] Refer to the attached Figure 2 As shown, an embodiment of the present application further provides a root cause location system for hydraulic pitch angle execution deviation, including: a data acquisition module 210 , a feature extraction module 220 and a root cause classification module 230 .

[0075] The data acquisition module 210 is used to collect the operating data of the wind turbine generator set in real time;

[0076] A feature extraction module 220 is used to perform target feature extraction based on the operating data, wherein the target feature extraction includes complex wind condition feature extraction, hydraulic drive system feature extraction, and blade coupled vibration feature extraction;

[0077] The root cause classification module 230 is used to perform root cause classification based on the extraction results of complex wind condition features, the extraction results of hydraulic drive system features, and the extraction results of blade coupled vibration features when a pitch tracking fault is triggered. The root cause classification includes complex wind conditions, hydraulic actuator abnormalities, blade coupled vibrations, and unclassified abnormalities.

[0078] The embodiments of the present application effectively improve diagnostic accuracy, significantly shorten single fault location time, effectively optimize operation and maintenance efficiency, and enhance system reliability by accurately distinguishing three root causes: pitch drive overrun caused by complex wind conditions, hydraulic actuator leakage / pressure anomaly, and angle oscillation caused by blade coupling vibration.

[0079] The present application also provides a computer-readable storage medium having a computer program stored thereon, wherein, when executed by a processor, the program implements the aforementioned method for locating the root cause of a hydraulic pitch angle execution deviation. For example, the method for locating the root cause of a hydraulic pitch angle execution deviation of the present application can be implemented via computer program instructions, and the relevant code can be stored in a computer-readable storage medium (such as a hard disk, SSD, or cloud server).

[0080] It is understood that the above embodiments are merely exemplary embodiments for illustrating the principles of the present application, and the present application is not limited thereto. Those skilled in the art may make various modifications and improvements without departing from the spirit and substance of the present application, and such modifications and improvements are also considered to be within the scope of protection of the present application.

Claims

1. A method for locating the root cause of hydraulic pitch angle execution deviation, characterized in that: include: Real-time collection of wind turbine operating data; Based on the operating data, target feature extraction is performed, wherein the target feature extraction includes complex wind condition feature extraction, hydraulic drive system feature extraction, and blade coupled vibration feature extraction; When a pitch tracking fault is triggered, root cause classification is performed based on the extraction results of the complex wind condition feature extraction, the extraction results of the hydraulic drive system feature extraction, and the extraction results of the blade coupled vibration feature extraction. The root cause classification includes complex wind conditions, hydraulic actuator abnormalities, blade coupled vibrations, and unclassified abnormalities.

2. The method according to claim 1, characterized in that The operation data includes: wind speed sampling information, pitch angle command value, actual execution value of pitch angle of three blades, hydraulic station pressure value and oil tank liquid level value, and performing target feature extraction based on the operation data includes: Calculating the average wind speed and turbulence intensity within a predetermined time window based on the wind speed sampling information to perform complex wind condition feature extraction; Calculating a pressure deviation between the real-time pressure of the hydraulic station and a preset pressure, and a level deviation between the real-time level of the oil tank and a preset minimum level, based on the hydraulic station pressure value and the oil tank liquid level value, so as to perform hydraulic drive system feature extraction; According to the pitch angle command value and the actual execution values ​​of the pitch angles of the three blades, it is respectively detected whether the actual execution values ​​of the pitch angles of the three blades oscillate around the pitch angle command value, and the fluctuation frequency characteristics of the blades are extracted during oscillation to perform the blade coupled vibration feature extraction.

3. The method according to claim 2, characterized in that Calculating the average wind speed and turbulence intensity within a predetermined time window based on the wind speed sampling information to perform complex wind condition feature extraction includes: Acquiring the wind speed sampling information within a continuous time window; The average wind speed and standard deviation of the wind speed sampling information are calculated, and the standard deviation is used as the turbulence intensity.

4. The method according to claim 3, characterized in that The calculation of the turbulence intensity within the predetermined time window can be expressed by the following formula: Wherein, Tur represents the turbulence intensity, V i Indicates the real-time wind speed sampling sequence, i represents the sequence number, V m represents the average wind speed, and 3000 is the number of sampling points.

5. The method according to claim 2, characterized in that When a pitch tracking fault is triggered, performing root cause classification based on the extraction results of the complex wind condition feature extraction, the extraction results of the hydraulic drive system feature extraction, and the extraction results of the blade coupled vibration feature extraction includes: If the average wind speed exceeds a first threshold and the turbulence intensity exceeds a second threshold, determining that the root cause is the complex wind condition; If the pressure deviation exceeds a third threshold or the liquid level deviation exceeds a fourth threshold, it is determined that the root cause is an abnormality of the hydraulic actuator; If the deviation between the extracted fluctuation frequency characteristic and the blade natural frequency characteristic is within a preset tolerance range, determining that the root cause is the blade coupled vibration; If none of the above conditions are met, it is determined that the root cause is the unclassified abnormality.

6. The method according to claim 2, characterized in that Before calculating the pressure deviation value between the real-time pressure of the hydraulic station and the preset design pressure, and the liquid level deviation value between the real-time liquid level of the oil tank and the preset minimum liquid level, the method includes: The hydraulic station pressure value and the oil tank liquid level value are subjected to low-pass filtering to filter out high-frequency noise signals.

7. The method according to claim 2, characterized in that The blade coupled vibration feature extraction includes: When it is detected that the actual execution value of the pitch angle of any blade continuously oscillates within a preset range around the pitch angle command value, performing Fourier transform on the oscillation signal; The transformed main frequency component is extracted as the fluctuation frequency feature.

8. The method according to claim 1, characterized in that The root cause classification is performed on the recorded data of a predetermined time period before the pitch tracking fault occurs, where the predetermined time period is 1 minute.

9. The method according to claim 1, characterized in that The preset tolerance range of the blade coupled vibration is ±5% of the blade natural frequency.

10. A root cause location system for hydraulic pitch angle execution deviation, characterized in that: include: Data acquisition module, used to collect real-time operating data of wind turbines; A feature extraction module, configured to perform target feature extraction based on the operating data, wherein the target feature extraction includes complex wind condition feature extraction, hydraulic drive system feature extraction, and blade coupled vibration feature extraction; A root cause classification module is used to perform root cause classification based on the extraction results of the complex wind condition feature extraction, the extraction results of the hydraulic drive system feature extraction, and the extraction results of the blade coupled vibration feature extraction when a pitch tracking fault is triggered. The root cause classification includes complex wind conditions, hydraulic actuator abnormalities, blade coupled vibrations, and unclassified abnormalities.

Images