A wind turbine yaw overspeed fault diagnosis method and early warning method

Abstract

This invention discloses a method for diagnosing and warning of yaw overspeed faults in wind turbines. The fault diagnosis method includes the following steps: 1) acquiring wind turbine operating data from a SCADA historical database; 2) identifying the yaw action process and yaw termination process in the wind turbine operating data to obtain yaw action data and yaw stationary data, and then filtering out yaw overspeed data from the yaw action data and yaw stationary data; 3) extracting wind turbine variables from the yaw overspeed data, obtaining variables characterizing yaw overspeed from the wind turbine variables, comparing the variables characterizing yaw overspeed with a preset threshold, and determining whether the wind turbine has experienced a yaw overspeed fault based on the comparison results. This invention has the advantages of accurate diagnosis and high reliability.

Description

Technical Field

[0001] This invention mainly relates to the field of wind power technology, specifically to a method for diagnosing and warning of yaw overspeed faults in wind turbine generators. Background Technology

[0002] In recent years, clean energy sources such as wind power, the third largest energy source after thermal and hydropower, have become the main force in improving the energy structure and protecting the environment, experiencing explosive growth over the past decade. The rapid upgrading and replacement of wind turbine models by various manufacturers, coupled with the significant human and material resources required for the operation and maintenance of wind turbines, necessitates continuous improvement in their reliability. Therefore, understanding the operating principles of wind turbines and analyzing common faults is of practical significance for preventing malfunctions and improving the operational stability of wind turbines.

[0003] Currently, the wind power industry has implemented various protective measures for severe yaw and overspeed conditions to prevent serious accidents such as wind turbine overruns and tower collapses. However, the identification of general overspeed conditions during wind turbine operation is not accurate or comprehensive enough, which can lead to a reduction in the service life of related equipment in the yaw system. Therefore, it is necessary to monitor and diagnose yaw and overspeed conditions of wind turbines, analyze the number of yaw and overspeed occurrences per unit time, issue fault warnings, and carry out timely maintenance to reduce the failure rate and improve power generation efficiency. Summary of the Invention

[0004] The technical problem to be solved by this invention is to provide a precise and reliable method for diagnosing and warning of yaw and overspeed faults in wind turbines, addressing the existing technical problems.

[0005] To solve the above-mentioned technical problems, the technical solution proposed by this invention is as follows:

[0006] A method for diagnosing yaw overspeed faults in wind turbine generators includes the following steps:

[0007] 1) Obtain wind turbine operation data from the SCADA historical database;

[0008] 2) Identify the yaw action process and yaw termination process in the wind turbine operation data to obtain yaw action data and yaw stationary data, and then filter out yaw overspeed data from the yaw action data and yaw stationary data.

[0009] 3) Extract wind turbine variables from the yaw overspeed data, then obtain variables representing yaw overspeed from the wind turbine variables, and then compare the variables representing yaw overspeed with preset thresholds. Based on the comparison results, determine whether the wind turbine has experienced a yaw overspeed fault.

[0010] Preferably, in step 2), each yaw action process and each yaw termination process in the wind turbine operation data are identified to obtain each yaw action, and then each yaw action is collected to obtain yaw action data; wherein each yaw action includes the yaw action process and the yaw termination process.

[0011] Preferably, in step 2), the yaw action process in the wind turbine operating data is identified based on the yaw brake pressure of the wind turbine.

[0012] Preferably, in step 2), the yaw action process in the wind turbine operating data is identified by the status of each yaw brake hydraulic valve.

[0013] Preferably, the yaw termination process in the wind turbine operating data is identified by analyzing the time distribution of the yaw brake pressure rising to the preset pressure during the yaw termination process.

[0014] Preferably, in step 3), the wind turbine variables include one or more of the following: wind speed, power, blade angle, generator speed, wind direction, yaw power, nacelle position, yaw brake pressure, yaw motor voltage, yaw motor current, yaw speed, master control status, and drive direction acceleration.

[0015] Preferably, the variables characterizing yaw overspeed include one or more of the following: cabin position, yaw speed, yaw motor current, and yaw motor voltage.

[0016] This invention also discloses a method for early warning of yaw overspeed faults in wind turbine generators, comprising:

[0017] Determine whether the wind turbine has yaw overspeed based on the wind turbine yaw overspeed fault diagnosis method described above; when yaw overspeed occurs, determine the level of yaw overspeed based on the yaw state, yaw speed and duration.

[0018] The severity of the yaw overspeed fault is determined based on the level of yaw overspeed and the number of times per unit time.

[0019] Corresponding fault warnings are issued based on the fault level of yaw and overspeed.

[0020] The present invention further discloses a computer-readable storage medium having a computer program stored thereon, the computer program executing the steps of the method described above when run by a processor.

[0021] The present invention also discloses a yaw overspeed fault diagnosis system for wind turbines, including a memory and a processor connected to each other. The memory stores a computer program, which executes the steps of the method described above when the processor runs the computer program.

[0022] Compared with the prior art, the advantages of the present invention are as follows:

[0023] The wind turbine yaw overspeed fault diagnosis method of the present invention uses the original SCADA data for subsequent yaw overspeed fault diagnosis, avoiding the problems of low feasibility and low accuracy of fault diagnosis caused by errors in single variable data, thereby improving the accuracy and reliability of diagnosis. By identifying the yaw action process and yaw termination process in the wind turbine operation data, and then constructing the corresponding yaw action data based on the corresponding identification results, both the yaw action data and the yaw stationary data are accurate and reliable, thereby further improving the accuracy and reliability of subsequent diagnosis. Attached Figure Description

[0024] Figure 1 This is a flowchart of an embodiment of the fault diagnosis method of the present invention.

[0025] Figure 2 This is a relationship diagram of the wind turbine yaw dataset of the present invention.

[0026] Figure 3 This is a flowchart of an embodiment of the fault diagnosis and early warning method of the present invention. Detailed Implementation

[0027] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0028] like Figure 1 As shown, the wind turbine yaw overspeed fault diagnosis method of this invention includes the following steps:

[0029] 1) Obtain wind turbine operation data from the SCADA historical database;

[0030] 2) Identify the yaw action process and yaw termination process in the wind turbine operation data, obtain yaw action data and yaw stationary data based on the identification results, and then filter out yaw overspeed data from the yaw action data and yaw stationary data; specifically, obtain yaw action data based on the identification results, and the remaining data in the wind turbine operation data is the yaw stationary data.

[0031] 3) Extract wind turbine variables from the yaw overspeed data, then obtain variables representing yaw overspeed from the wind turbine variables, and then compare the variables representing yaw overspeed with preset thresholds. Based on the comparison results, determine whether the wind turbine has experienced a yaw overspeed fault.

[0032] The wind turbine yaw overspeed fault diagnosis method of the present invention uses raw SCADA data for subsequent yaw overspeed fault diagnosis, avoiding the problems of low feasibility and low accuracy of fault diagnosis caused by errors in single variable data, thereby improving the accuracy and reliability of diagnosis. By identifying the yaw action process and yaw termination process in the wind turbine operation data, and then constructing corresponding yaw action data based on the corresponding identification results, the yaw action data and yaw stationary data are accurate and reliable, thereby further improving the accuracy and reliability of subsequent diagnosis.

[0033] In one specific embodiment, in step 2), each yaw action process and each yaw termination process in the wind turbine operating data are identified to obtain each yaw action. Then, the yaw actions are aggregated to obtain yaw action data. Each yaw action includes a yaw action process and a yaw termination process. By identifying the yaw action process and the yaw termination process from these two operating conditions, and then combining the corresponding yaw action process and the yaw termination process to obtain a complete yaw action, the accuracy and reliability of the yaw action data are ensured, which is beneficial for subsequent fault analysis.

[0034] Specifically, in step 2), the yaw action process in the wind turbine operating data is identified based on the yaw brake pressure. If the yaw brake pressure value is lower than 20 bar (this value is selected based on the wind turbine brake pressure design value), it indicates that the wind turbine is in a half-brake state and is in the yaw action process.

[0035] Alternatively, the yaw action process in the wind turbine's operating data can be identified based on the status of each yaw brake hydraulic valve. For example, when the wind turbine's yaw brake hydraulic valve is open (i.e., the value of yaw brake hydraulic valve 1 is 1 and the value of yaw brake hydraulic valve 2 is 1, where 1 represents open), it is in the yaw action process.

[0036] In practical applications, one of the above-mentioned suitable identification methods can be selected to identify the yaw process. Of course, the identification results of two methods can also be combined to finally determine whether a yaw process has occurred. By comprehensively judging multiple identification results (such as if both identification results indicate a yaw process, then it can be determined to be a yaw process), the misjudgment of a single method can be avoided, and the accuracy of subsequent diagnosis can be improved.

[0037] Specifically, in step 2), the yaw termination process in the wind turbine operating data is identified by analyzing the time distribution of the yaw brake pressure rise to the preset pressure during the yaw termination process. Specifically, this includes two methods:

[0038] 1. Analyze the time distribution of yaw brake pressure rise to 120 bar during the yaw termination process. Combined with the operating principle of wind turbine units (the time distribution of pressure rise varies for different wind turbines. Generally speaking, the time distribution of pressure rise to 120 bar from the end moment is more than 98% in the 5s-7s range, which can be regarded as the unit being in a completely stationary state), determine the yaw termination process.

[0039] 2. Analyze the closing time of the yaw brake hydraulic valve of the wind turbine (i.e. the moment when the values ​​of yaw brake hydraulic valves 1 and 2 jump from 1 to 0, where 1 represents opening and 0 represents closing), the closing time of the yaw brake hydraulic valve and 6 seconds thereafter (design threshold, which can be selected according to actual conditions), and determine the end of the yaw process.

[0040] In practical applications, one of the above-mentioned suitable identification methods can be selected to identify the yaw termination process. Of course, the identification results of two methods can also be combined to finally determine whether a yaw termination process has occurred. By comprehensively judging multiple identification results (such as if both identification results indicate a yaw termination process, then it can be determined to be a yaw termination process), the misjudgment of a single method can be avoided, and the accuracy of subsequent diagnosis can be improved.

[0041] Based on the operating principle of wind turbines, this invention identifies two operating conditions: the yaw action process and the yaw termination process. It then constructs complete yaw action data and obtains yaw stationary data. Based on the yaw action data and yaw stationary data, it diagnoses and warns of yaw overspeed faults, thereby improving the accuracy of subsequent yaw overspeed diagnosis and facilitating the analysis of the causes of the faults.

[0042] like Figure 3 As shown, the present invention also discloses a method for early warning of yaw overspeed faults in wind turbine generators, comprising:

[0043] Determine whether the wind turbine has experienced yaw overspeed based on the yaw overspeed fault diagnosis method described above; when yaw overspeed occurs, determine the level of yaw overspeed based on the yaw state, yaw speed, and duration of the wind turbine.

[0044] The severity of the yaw overspeed fault is determined based on the level of yaw overspeed and the number of times per unit time.

[0045] Corresponding fault warnings are issued based on the fault level of yaw and overspeed.

[0046] The wind turbine yaw overspeed fault early warning method of the present invention determines the level of yaw overspeed based on the yaw state, yaw speed and duration of the wind turbine, and then determines the fault level of yaw overspeed based on the level of yaw overspeed and the number of times per unit time; and then performs corresponding fault early warning based on the fault level of yaw overspeed, thereby improving the precision of the early warning.

[0047] This invention further discloses a computer-readable storage medium storing a computer program thereon, which, when run by a processor, performs the steps of the method described above. This invention also discloses a wind turbine yaw overspeed fault diagnosis system, comprising a memory and a processor interconnected, wherein the memory stores a computer program, which, when run by a processor, performs the steps of the method described above. The medium and system of this invention, corresponding to the above-described diagnostic methods, also possess the advantages described above.

[0048] To better understand the above technical solution, the following will provide a detailed description of the technical solution in conjunction with the accompanying drawings and specific implementation methods:

[0049] (1) Acquisition and preprocessing of SCADA operation data at the second level

[0050] Wind turbine operation data is obtained from the SCADA historical database, with a data granularity of 1 second. The acquired second-level data is preprocessed—abnormal, invalid, and duplicate data are removed (e.g., differential processing of nacelle position and differential sliding window processing are performed)—to construct a new feature dataset, denoted as S. origin .

[0051] Of course, in other embodiments, feature extraction can also be performed based on minute-level data: the preprocessed second-level operation data can be aggregated to obtain 1-minute granular wind turbine operation data, and the status labels of each 1-minute granular operation data can be determined, etc.

[0052] (2) Screening for yaw and overspeed conditions

[0053] Based on the operating principle of wind turbine units, from dataset S origin The process of each yaw action and the end of each yaw action of the wind turbine can be identified.

[0054] The yaw action process is identified in the following two ways:

[0055] 1. Identification is based on the threshold designed for the yaw brake pressure of the wind turbine: when the wind turbine is in a half-brake state (yaw brake pressure value is below 20 bar), it is in the process of yaw action;

[0056] 2. Identification is made by the design values ​​of yaw brake hydraulic valve 1 and yaw brake hydraulic valve 2: When the yaw brake hydraulic valve of the wind turbine is open (i.e., the value of yaw brake hydraulic valve 1 is 1 and the value of yaw brake hydraulic valve 2 is 1, where 1 represents open), it is in the yaw action process.

[0057] In practical applications, a suitable identification method can be chosen to identify the yaw motion process. Alternatively, the results of two identification methods can be combined to ultimately determine whether a yaw motion process has occurred. If both identification results indicate a yaw motion process, then it can be determined to be a yaw motion process, improving the accuracy of the identification.

[0058] The yaw termination process is identified in the following two ways:

[0059] 1. Analyze the time distribution of yaw brake pressure rise to 120 bar during the yaw termination process. Combined with the operating principle of wind turbine units (the time distribution of pressure rise varies for different wind turbines. Generally speaking, the time distribution of pressure rise to 120 bar from the end moment is more than 98% in the 5s-7s range, which can be regarded as the unit being in a completely stationary state), determine the yaw termination process.

[0060] 2. Analyze the closing time of the yaw brake hydraulic valve of the wind turbine (i.e. the moment when the values ​​of yaw brake hydraulic valves 1 and 2 jump from 1 to 0, where 1 represents opening and 0 represents closing), the closing time of the yaw brake hydraulic valve and 6 seconds thereafter (design threshold, which can be selected according to actual conditions), and determine the end of the yaw process.

[0061] Similarly, in practical applications, a suitable identification method can be chosen to identify the yaw termination process. Alternatively, the results of both methods can be combined to ultimately determine whether a yaw termination process has occurred. If both identification results indicate a yaw termination process, then it can be determined to be a yaw termination process, thus improving the accuracy of the identification.

[0062] Depending on the aircraft type, a suitable identification method is selected for both the yaw maneuver process and the yaw termination process to constitute a single complete yaw maneuver, denoted as S. yaw_i (i = 1, 2, ..., n); the set of all yaw actions is denoted as the complete yaw action data S. yaw ,but From S origin The dataset after removing all complete yaw motion data is the set of yaw stationary data, denoted as S. yaws There is S yaws =S origin -S yaw The wind turbine operating data can be broken down into: complete yaw action data (yaw action process + yaw termination process) plus yaw stationary data.

[0063] By characterizing key variables of yaw overspeed (nacelle position, yaw speed, yaw motor current, yaw motor voltage, etc.), and based on the wind turbine operating principle and the protection threshold design of the wind turbine yaw system, the S... yaw and S yaws Perform the following operations:

[0064] 1. From S yaw Select data segments for yaw and overspeed conditions, and denote each yaw and overspeed condition as S. yaw_overspd_i (i = 1, 2, ..., n), the set of all yaw and overspeed conditions is denoted as S. yaw_overspd Then S yaw_overspd ={S yaw_overspd_1 S yaw_overspd_2 S yaw_overspd_n}

[0065] 2. From S yaws Filter the data segments of yaw and speeding, and record each instance of yaw and speeding as S. yaws_overspd_i (i = 1, 2, ..., n), the set of each yaw overspeed is denoted as S. yaws_overspd Then S yaws_overspd ={S yaws_overspd_1 S yaws_overspd_2 S yaws_overspd_n}

[0066] The relationships between the above datasets are as follows: Figure 2 As shown.

[0067] (3) Diagnosis and early warning of yaw and overspeed faults

[0068] After filtering out the yaw overspeed data of the wind turbine, the yaw overspeed fault of the wind turbine is diagnosed and an early warning is issued.

[0069] The diagnostic methods are briefly described below:

[0070] 1. For each speeding data point, slide a window for 120 seconds to select a data snapshot;

[0071] 2. Extract relevant variables from the data snapshot: wind speed, power, blade angle, generator speed, wind direction, yaw power, nacelle position, yaw brake pressure, yaw motor voltage, yaw motor current, yaw speed, master control status and drive direction acceleration, etc., and draw a time series diagram;

[0072] 3. Based on the operating principle of the wind turbine, analyze the state of the wind turbine under this operating condition by characterizing key variables of yaw overspeed, such as nacelle position, yaw speed, yaw motor current, and yaw motor voltage, to diagnose whether the unit has actually experienced overspeed. Specifically, for example:

[0073] Overspeed level 1: In the yaw maneuver state (corresponding to the yaw maneuver process + yaw end process), the yaw speed is >0.6° / s, lasting for 2 seconds;

[0074] Overspeed level 2: In the yaw maneuver state (corresponding to the yaw maneuver process + yaw end process), the yaw speed is >1° / s, lasting for 3 seconds;

[0075] Overspeed level 3: Under all operating conditions (corresponding to the yaw action process + yaw end process + yaw stationary condition), the yaw speed is >1.5° / s, lasting for 3 seconds.

[0076] The thresholds for yaw speed and duration mentioned above should be selected based on the actual situation.

[0077] After diagnosing a wind turbine overspeed, the yaw overspeed screening model, based on thresholds set according to the wind turbine's operating principles, accurately indicates that the wind turbine is experiencing severe operating conditions. Based on the number of yaw overspeed occurrences per unit time, timely fault warnings are issued. Figure 3 As shown. The specific early warning method is as follows:

[0078] Fault Level 1: Within a unit of time, overspeed Level 1 conditions occurred 5 times, overspeed Level 2 conditions occurred 0 times, and overspeed Level 3 conditions occurred 0 times.

[0079] Fault Level 2: Within a unit of time, the overspeed level 2 condition occurs 2 times and the overspeed level 3 condition occurs 0 times.

[0080] Fault Level 3: Overspeed Level 3 occurs once per unit time.

[0081] When the overspeed condition reaches fault level 1, 2 and 3, a corresponding fault warning will be given.

[0082] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A method for diagnosing yaw overspeed faults in wind turbine generators, characterized in that, Including the following steps: 1) Obtain wind turbine operation data from the SCADA historical database; 2) Identify the yaw action process and yaw termination process in the wind turbine operation data, obtain yaw action data and yaw stationary data based on the identification results, and then filter out yaw overspeed data from the yaw action data and yaw stationary data; specifically, obtain the yaw action data based on the identification results, and the remaining data in the wind turbine operation data is the yaw stationary data. 3) Extract wind turbine variables from yaw overspeed data, then obtain variables representing yaw overspeed from wind turbine variables, and then compare the variables representing yaw overspeed with preset thresholds. Based on the comparison results, determine whether the wind turbine has a yaw overspeed fault. In step 2), the yaw action process in the wind turbine operating data is identified based on the yaw brake pressure; or the yaw action process in the wind turbine operating data is identified based on the status of each yaw brake hydraulic valve. The yaw termination process in wind turbine operating data is identified by analyzing the time distribution of the yaw brake pressure rise to the preset pressure during the yaw termination process. In step 2), each yaw action process and each yaw termination process in the wind turbine operation data are identified to obtain each yaw action, and then each yaw action is collected to obtain yaw action data; wherein each yaw action includes the yaw action process and the yaw termination process.

2. The method for diagnosing yaw overspeed faults in wind turbine units according to claim 1, characterized in that, In step 3), the wind turbine variables include one or more of the following: wind speed, power, blade angle, generator speed, wind direction, yaw power, nacelle position, yaw brake pressure, yaw motor voltage, yaw motor current, yaw speed, master control status, and drive direction acceleration.

3. The wind turbine yaw overspeed fault diagnosis method according to claim 2, characterized in that, The variables characterizing yaw overspeed include one or more of the following: cabin position, yaw speed, yaw motor current, and yaw motor voltage.

4. A method for early warning of yaw overspeed fault in wind turbine generators, characterized in that, include: The wind turbine yaw overspeed fault diagnosis method according to any one of claims 1-3 determines whether the wind turbine has yaw overspeed. When yaw and speeding occur, the level of yaw and speeding is determined based on the yaw state, yaw speed, and duration. The severity of the yaw overspeed fault is determined based on the level of yaw overspeed and the number of times per unit time. Corresponding fault warnings are issued based on the fault level of yaw and overspeed.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program, when run by a processor, performs the steps of the method as described in any one of claims 1 to 3.

6. A wind turbine yaw overspeed fault diagnosis system, comprising a memory and a processor interconnected, wherein the memory stores a computer program, characterized in that, The computer program, when run by a processor, performs the steps of the method as described in any one of claims 1 to 3.

Images