A wind turbine blade fault discrimination method based on standard deviation analysis of variable pitch signals

Abstract

The application discloses a wind turbine blade fault discrimination method based on standard deviation analysis of a variable pitch signal, comprising the following steps: 1) generating a function block based on the IEC61311-3 international standard, which can be called by a PLC; 2) taking actual measurement data of a variable pitch system as input signals of the carrier, and the signals have measurement noise interference, so that a sliding filter algorithm is used to process actual measurement values of the pitch angles of three blades; 3) after the sliding filter algorithm is used to process the actual measurement values of the pitch angles of the variable pitch system, pitch angle values of the three blades after the sliding filter algorithm are obtained, and standard deviations of the pitch angle values of the three blades within 60 ms after the sliding filter algorithm are calculated; and 4) if the absolute value of the difference between the standard deviations of the pitch angles of any two blades and the ratio of the standard deviations of the two blades differ by more than 10%, it is judged that the blades have faults, otherwise, the blades have no faults. The application realizes wind turbine blade fault discrimination.

Description

Technical Field

[0001] This invention belongs to the field of wind power generation technology, specifically relating to a method for identifying wind turbine blade faults based on pitch signal standard deviation analysis. Background Technology

[0002] The trend towards larger wind turbines is becoming increasingly evident. The complex aeroelastic characteristics of large, long, flexible blades make it difficult to effectively estimate their displacement response during actual operation using commercial software. Furthermore, defects in large, long, flexible blades caused by manufacturing processes that do not meet design requirements or substandard raw materials are difficult to detect effectively. In recent years, an increasing number of wind turbines have experienced serious accidents such as blade sweeping and turbine collapse after commissioning. Blade defects are a major cause of sweeping accidents.

[0003] To enable wind turbines to detect potential fault signals at an early stage of blade failure, the main control program needs to use advanced algorithms to process the turbine's operating data and obtain potential fault signals. However, regarding blade fault identification, most main control programs still have the following major shortcomings in judging blade faults:

[0004] (1) Most master control programs only judge blade faults by judging the asynchronous pitch angles of the three blades, and only make simple difference comparisons when the blade pitch angles are asynchronous. This method has poor reliability. When the entire wind turbine vibrates significantly due to a single blade fault, it is impossible to make timely alarms by relying solely on the asynchronous pitch angles.

[0005] (2) Some OEMs also equip their blades with online vibration monitoring systems, but these systems are independent of the main control program of the wind turbine. The blade fault information fed back by the online vibration monitoring system is often delayed due to human factors.

[0006] (3) Some large wind turbine units that have been newly installed in recent years have not been equipped with blade online vibration monitoring systems;

[0007] (4) Blade online vibration monitoring systems are expensive and require separate equipment to be installed inside the wind turbine hub;

[0008] (5) Some units that have installed blade online vibration monitoring systems still experienced blade sweeping and even tower collapse accidents. Summary of the Invention

[0009] This invention provides a wind turbine blade fault identification method based on pitch signal standard deviation analysis, which is easy to design and can effectively detect wind turbine blade faults and pitch actuator faults.

[0010] This invention is achieved through the following technical solution:

[0011] A method for identifying wind turbine blade faults based on pitch signal standard deviation analysis includes the following steps:

[0012] 1) Based on the IEC61311-3 international standard, function blocks that can be called by the PLC serve as the carrier;

[0013] 2) The actual measurement data of the pitch system is used as the input signal of the carrier. This signal is subject to measurement noise interference. Therefore, the measured values ​​of the pitch angle of the three blades are processed by a sliding filter algorithm. The time period of the sliding filter algorithm is set to 60ms. That is, when the task period of the main control PLC is 10ms, the measurement values ​​of 6 periods are processed by sliding filter. When the task period is 20ms, the measurement values ​​of 3 periods are processed by sliding filter.

[0014] 3) Based on the sliding filter processing of the measured pitch angle signal of the pitch system, the pitch angle values ​​of the three blades after sliding filter are obtained respectively, and the standard deviation of the pitch angle of the three blades after sliding filter within 60ms is calculated respectively.

[0015] 4) If the absolute value of the difference between the standard deviations of the pitch angles of any two blades and the ratio of the standard deviations of the two blades differ by more than 10%, then the blade is judged to have a fault; otherwise, the blade does not have a fault.

[0016] A further improvement of this invention is that, for the acquisition of pitch system signals, only the pitch angle information measured in real time by the PLC is acquired, and this signal is refreshed once every 10ms.

[0017] A further improvement of this invention is that the 60ms sampling duration is determined for units with a capacity of 2.0MW or higher.

[0018] A further improvement of this invention is that, in current high-capacity advanced megawatt-class wind turbines, there are two ways to measure the pitch angle: using an A encoder located at the blade root and a B encoder located at the pitch drive. The B encoder has higher measurement accuracy than the A encoder, so the pitch angle measurement value of the pitch system is measured using the B encoder value.

[0019] A further improvement of this invention is that, for the calculation of the standard deviation of the blade pitch angle measurement, a sampling time of 60ms is determined, and the standard deviation is calculated once every 10ms refresh. That is, the blade pitch angle measurement and the calculation of the blade pitch angle standard deviation are performed synchronously.

[0020] A further improvement of this invention lies in the fact that, when determining blade faults based on the calculation results of the blade pitch angle standard deviation:

[0021] Record separately:

[0022] The standard deviation value A is: the absolute value of the difference between the standard deviation of the pitch angle measurement of the first blade and the standard deviation of the pitch angle measurement of the second blade, and the ratio of the standard deviation of the pitch angle measurement of the first blade or the standard deviation of the pitch angle measurement of the second blade, where the larger value is taken.

[0023] The standard deviation value B is: the absolute value of the difference between the standard deviation of the second blade's pitch angle measurement and the standard deviation of the second blade's pitch angle measurement, and the ratio of the standard deviation of the second blade's pitch angle measurement or the standard deviation of the third blade's pitch angle measurement, where the larger value is taken.

[0024] The standard deviation C is the ratio of the absolute value of the difference between the standard deviation of the third blade's pitch angle measurement and the standard deviation of the first blade's pitch angle measurement to either the standard deviation of the third blade's pitch angle measurement or the standard deviation of the first blade's pitch angle measurement, where the larger value is taken.

[0025] If the standard deviation A is greater than or equal to 10%, the standard deviation B is greater than or equal to 10%, and the standard deviation C is less than 10%, then the second blade is judged to be faulty.

[0026] A further improvement of the present invention is that if the standard deviation value B is greater than or equal to 10%, and the standard deviation value C is greater than or equal to 10%, and the standard deviation value A is less than 10%, then the third blade is judged to be faulty.

[0027] A further improvement of the present invention is that if the standard deviation value C is greater than or equal to 10%, and the standard deviation value A is greater than or equal to 10%, and the standard deviation value B is less than 10%, then the first blade is judged to have a fault.

[0028] The present invention has at least the following beneficial technical effects:

[0029] This invention uses function blocks conforming to the IEC61311-3 international standard as a carrier. Based on the actual pitch angle of the three blades collected by the main control PLC, and after filtering and algorithm processing by the function blocks, the fault deviation of each blade can be accurately obtained, realizing the fault identification of wind turbine blades. It overcomes the high risk and inaccuracy of traditional wind turbine blade identification methods, such as identifying potential blade faults based on online blade monitoring systems and visual inspection of blade faults by maintenance personnel using telescopes and suspended platforms. It has the following advantages:

[0030] First: The algorithm based on function blocks can be embedded in the main control PLC without occupying or adding any mechanical structure;

[0031] Second: Compared with the existing methods in the industry that use telescopes and suspended platforms for visual inspection by maintenance personnel, this method does not pose any personal danger and is safe and reliable;

[0032] Third: The filtering algorithm for the actual measured blade pitch angle involved in this invention has the advantages of being mature, reliable, and having low computational load, and can effectively remove spikes in the blade pitch angle measurement signal.

[0033] Fourth: The determination of the sampling period and sampling time of the three blade pitch angles involved in this invention is based on numerous wind turbine blade failure accident analysis cases and has a rich engineering practice foundation.

[0034] Fifth: Based on the ideas and algorithms of this invention, the status of the turbine blades can be evaluated offline according to the actual operating data of the wind turbine. This method is highly practical and reliable, and can save a lot of wind farm operation and maintenance costs. Attached Figure Description

[0035] Figure 1 This is a flowchart of a wind turbine blade fault identification method based on pitch signal standard deviation analysis according to the present invention.

[0036] Figure 2 This is a schematic diagram of the core steps of a wind turbine blade fault identification method based on pitch signal standard deviation analysis according to the present invention.

[0037] Figure 3 This is a schematic diagram illustrating the actual measured values ​​of the pitch angles of the three blades before and after a broken blade, based on the standard deviation analysis of the pitch signal, according to the present invention.

[0038] Figure 4 This is a schematic diagram of the actual measured values ​​of the pitch angle of the three blades before and after a broken blade, based on the standard deviation analysis of the pitch signal, after being processed by sliding filtering.

[0039] Figure 5 This is a schematic diagram comparing the standard deviations of the pitch angles of the three blades of a broken blade before and after the breakage, based on the standard deviation analysis of the pitch signal. Detailed Implementation

[0040] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0041] like Figure 1 As shown, the present invention provides a method for wind turbine blade fault identification based on pitch signal standard deviation analysis, comprising the following steps:

[0042] 1) Based on the IEC61311-3 international standard, function blocks that can be called by the PLC serve as the carrier;

[0043] 2) The actual measurement data of the pitch system is used as the input signal of the carrier. This signal is subject to measurement noise interference. Therefore, the measured values ​​of the pitch angle of the three blades are processed by a sliding filter algorithm. The time period of the sliding filter algorithm is set to 60ms. That is, when the task period of the main control PLC is 10ms, the measurement values ​​of 6 periods are processed by sliding filter. When the task period is 20ms, the measurement values ​​of 3 periods are processed by sliding filter.

[0044] 3) Based on the sliding filter processing of the measured pitch angle signal of the pitch system, the pitch angle values ​​of the three blades after sliding filter are obtained respectively, and the standard deviation of the pitch angle of the three blades after sliding filter within 60ms is calculated respectively.

[0045] 4) If the absolute value of the difference between the standard deviations of the pitch angles of any two blades and the ratio of the standard deviations of the two blades differ by more than 10%, then the blade is judged to have a fault; otherwise, the blade does not have a fault.

[0046] 01: For the acquisition of pitch system signals, only the pitch angle information measured in real time by the PLC is collected. This signal is refreshed once every 10ms, and the sampling accuracy is high.

[0047] 02: For the actual measurement signals of the pitch system, there are many filtering functions that can be used. This invention adopts the method of calculating the average value of sliding filter, which has low computational complexity and good effect.

[0048] 03: The calculation results will have different deviations depending on the sampling duration. This invention determines a sampling duration of 60ms for units with a capacity of 2.0MW or above.

[0049] 04: Currently, there are generally two ways to measure the pitch angle of large-capacity advanced megawatt-level wind turbines: one is to use an A encoder located at the blade root, and the other is to use a B encoder located at the pitch drive. The B encoder has higher measurement accuracy than the A encoder. Therefore, the pitch angle measurement value of the pitch system is measured by the B encoder.

[0050] 05: For the calculation of the standard deviation of the blade pitch angle measurement, a sampling time of 60ms is determined, and the standard deviation is calculated once every 10ms refresh of the program. That is, the blade pitch angle measurement and the calculation of the standard deviation of the blade pitch angle are performed synchronously.

[0051] 06: Based on the calculation results of the standard deviation of the blade pitch angle, the blade fault can be identified. The more detailed steps are as follows:

[0052] Step 1: Record the following separately:

[0053] The standard deviation value A is: the absolute value of the difference between the standard deviation of the pitch angle measurement of the first blade and the standard deviation of the pitch angle measurement of the second blade, and the ratio of the standard deviation of the pitch angle measurement of the first blade or the standard deviation of the pitch angle measurement of the second blade, where the larger value is taken.

[0054] The standard deviation value B is: the absolute value of the difference between the standard deviation of the second blade's pitch angle measurement and the standard deviation of the second blade's pitch angle measurement, and the ratio of the standard deviation of the second blade's pitch angle measurement or the standard deviation of the third blade's pitch angle measurement, where the larger value is taken.

[0055] The standard deviation C is the ratio of the absolute value of the difference between the standard deviation of the third blade's pitch angle measurement and the standard deviation of the first blade's pitch angle measurement to either the standard deviation of the third blade's pitch angle measurement or the standard deviation of the first blade's pitch angle measurement, where the larger value is taken.

[0056] Step Two:

[0057] If the standard deviation A is greater than or equal to 10%, the standard deviation B is greater than or equal to 10%, and the standard deviation C is less than 10%, then it can be determined that the second blade is faulty.

[0058] Step 3: If the standard deviation value B is greater than or equal to 10%, the standard deviation value C is greater than or equal to 10%, and the standard deviation value A is less than 10%, then it can be determined that the third blade is faulty.

[0059] Step 4: If the standard deviation C is greater than or equal to 10%, the standard deviation A is greater than or equal to 10%, and the standard deviation B is less than 10%, then it can be determined that the first blade has malfunctioned.

[0060] Example

[0061] In 2021, a blade fracture accident occurred in a 3MW onshore generator unit. The operating data after the blade failure was used for example analysis.

[0062] Step 1: Measure the blade pitch angle of a 3.2MW unit and obtain the actual pitch angle of the three blades, as shown in Figure 3.

[0063] Step 2: Apply a sliding filter to the pitch angle of the three blades to remove corresponding burrs, such as... Figure 4 As shown:

[0064] Step 3: Calculate the standard deviation of the pitch angle for each of the three blades, and obtain three standard deviation values, where:

[0065] The standard deviation value A is: the absolute value of the difference between the standard deviation of the pitch angle measurement of the first blade and the standard deviation of the pitch angle measurement of the second blade, and the ratio of the standard deviation of the pitch angle measurement of the first blade or the standard deviation of the pitch angle measurement of the second blade, where the larger value is taken.

[0066] The standard deviation value B is: the absolute value of the difference between the standard deviation of the second blade's pitch angle measurement and the standard deviation of the third blade's pitch angle measurement, and the ratio of the standard deviation of the second blade's pitch angle measurement or the standard deviation of the third blade's pitch angle measurement, where the larger value is taken.

[0067] The standard deviation C is the ratio of the absolute value of the difference between the standard deviation of the third blade's pitch angle measurement and the standard deviation of the first blade's pitch angle measurement to either the standard deviation of the third blade's pitch angle measurement or the standard deviation of the first blade's pitch angle measurement, where the larger value is taken.

[0068] Obtain, such as Figure 5 As shown in Table 1:

[0069] Table 1 shows the actual data comparing the standard deviations of the pitch angles of the three blades before and after a broken blade, based on the standard deviation analysis of the pitch signal according to the present invention.

[0070]

[0071]

[0072] In Table 1, "cycle" represents the task cycle of the wind turbine main controller.

[0073] When cycle is 0, it indicates the moment when the blade failure occurred;

[0074] When cycle is less than 0, it indicates that the blade has not yet failed.

[0075] When cycle is greater than 0, it indicates that a blade failure has occurred.

[0076] Based on the case of wind turbine blade breakage accidents, it can be determined that blade #3 has malfunctioned.

[0077] The present invention describes a method for identifying blade faults based on the standard deviation analysis of pitch signals.

[0078] Step 1: The following three variables,

[0079] Pitch angle measurement value 1;

[0080] Pitch angle measurement value 2;

[0081] Pitch angle measurement value 3;

[0082] These are the measured pitch angle values ​​for the first to third blades, respectively.

[0083] Step 2: Perform sliding filtering on the acquired pitch angle measurements. The sliding filtering time is specified as 60ms. That is, when the main control PLC's task cycle is 10ms, sliding filtering is performed on the measurements from 6 cycles; when the task cycle is 20ms, sliding filtering is performed on the measurements from 3 cycles.

[0084] Step 3: Record the average value of the obtained pitch angle filter values ​​as follows:

[0085] Pitch angle sliding filter value 1;

[0086] Pitch angle sliding filter value 2;

[0087] Pitch angle sliding filter value 3;

[0088] Step 4: Calculate the standard deviation of the pitch angle for each blade over 60ms. The standard deviation of the pitch angle for each blade is denoted as follows:

[0089] Standard deviation of the first blade pitch angle measurement;

[0090] Standard deviation of the second blade pitch angle measurement;

[0091] Standard deviation of the third blade pitch angle measurement;

[0092] Based on the obtained standard deviation of the pitch angle for each blade, the deviation value of the standard deviation is calculated separately:

[0093] The standard deviation value A is: the absolute value of the difference between the standard deviation of the first blade pitch angle measurement and the standard deviation of the second blade pitch angle measurement, and the ratio of the standard deviation of the first blade pitch angle measurement or the standard deviation of the second blade pitch angle measurement, where the larger value is taken.

[0094] The standard deviation value B is: the absolute value of the difference between the standard deviation of the second blade pitch angle measurement and the standard deviation of the second blade pitch angle measurement, and the ratio of the standard deviation of the second blade pitch angle measurement or the standard deviation of the third blade pitch angle measurement, where the larger value is taken.

[0095] The standard deviation C is the ratio of the absolute value of the difference between the standard deviation of the third blade pitch angle measurement and the standard deviation of the first blade pitch angle measurement to the standard deviation of either the third blade pitch angle measurement or the first blade pitch angle measurement, where the larger value is taken.

[0096] Step 5: If the following situation occurs, the main controller will report that the blade has malfunctioned; otherwise, no fault will be reported.

[0097] If the standard deviation A is greater than or equal to 10%, the standard deviation B is greater than or equal to 10%, and the standard deviation C is less than 10%, then it can be determined that the second blade is faulty.

[0098] If the standard deviation B is greater than or equal to 10%, the standard deviation C is greater than or equal to 10%, and the standard deviation A is less than 10%, then it can be determined that the third blade is faulty.

[0099] If the standard deviation C is greater than or equal to 10%, the standard deviation A is greater than or equal to 10%, and the standard deviation B is less than 10%, then it can be determined that the first blade has malfunctioned.

[0100] Although the present invention has been described in detail above with general descriptions and specific embodiments, modifications or improvements can be made to it, which will be obvious to those skilled in the art. Therefore, all such modifications or improvements made without departing from the spirit of the present invention fall within the scope of protection claimed by the present invention.

Claims

1. A method for identifying wind turbine blade faults based on pitch signal standard deviation analysis, characterized in that, Includes the following steps: 1) Function blocks generated based on the IEC61311-3 international standard and callable by the PLC serve as the carrier; 2) The actual measurement data of the pitch system is used as the input signal of the carrier. This signal is subject to measurement noise interference. Therefore, the measured values ​​of the pitch angle of the three blades are processed by a sliding filter algorithm. The time period of the sliding filter algorithm is set to 60ms. That is, when the task period of the main control PLC is 10ms, the measured values ​​of 6 periods are processed by sliding filter. When the task period is 20ms, the measured values ​​of 3 periods are processed by sliding filter. 3) Based on the sliding filter processing of the measured pitch angle signal of the pitch system, the pitch angle values ​​of the three blades after sliding filter are obtained respectively. The standard deviation of the pitch angle of these three blades after sliding filter within 60ms is calculated respectively. 4) If the absolute value of the difference in the standard deviation of the pitch angle between any two blades differs from the ratio of the standard deviation of the two blades by more than 10%, then the blade is considered to have a fault; otherwise, the blade is not faulty. To calculate the standard deviation of the blade pitch angle measurement, a sampling time of 60ms is determined, and the standard deviation is calculated once every 10ms refresh. That is, the blade pitch angle measurement and the calculation of the blade pitch angle standard deviation are performed synchronously. When using the calculation results of the standard deviation of blade pitch angle to determine blade faults: Record separately: The standard deviation value A is: the absolute value of the difference between the standard deviation of the first blade's pitch angle measurement and the standard deviation of the second blade's pitch angle measurement, and the ratio of the standard deviation of the first blade's pitch angle measurement or the standard deviation of the second blade's pitch angle measurement, where the standard deviation value A is the larger value. The standard deviation value B is: the absolute value of the difference between the standard deviation of the second blade's pitch angle measurement and the standard deviation of the second blade's pitch angle measurement, and the ratio of the standard deviation of the second blade's pitch angle measurement or the standard deviation of the third blade's pitch angle measurement, where the standard deviation value B is the larger value. The standard deviation C is the ratio of the absolute value of the difference between the standard deviation of the third blade's pitch angle measurement and the standard deviation of the first blade's pitch angle measurement to either the standard deviation of the third blade's pitch angle measurement or the standard deviation of the first blade's pitch angle measurement, where the larger value of the standard deviation C is taken. If the standard deviation A is greater than or equal to 10%, the standard deviation B is greater than or equal to 10%, and the standard deviation C is less than 10%, then the second blade is judged to be faulty.

2. The method for wind turbine blade fault identification based on pitch signal standard deviation analysis according to claim 1, characterized in that, For the acquisition of pitch system signals, only the pitch angle information measured in real time by the PLC is collected, and this signal is refreshed once every 10ms.

3. The method for wind turbine blade fault identification based on pitch signal standard deviation analysis according to claim 1, characterized in that, The 60ms sampling duration is determined for units with a capacity of 2.0MW or higher.

4. The method for wind turbine blade fault identification based on pitch signal standard deviation analysis according to claim 1, characterized in that, If the standard deviation B is greater than or equal to 10%, the standard deviation C is greater than or equal to 10%, and the standard deviation A is less than 10%, then the third blade is judged to be faulty.

5. The method for wind turbine blade fault identification based on pitch signal standard deviation analysis according to claim 4, characterized in that, If the standard deviation C is greater than or equal to 10%, the standard deviation A is greater than or equal to 10%, and the standard deviation B is less than 10%, then the first blade is judged to be faulty.

Images