Wind turbine generator fault early warning method and system based on sudden change characterization value

Abstract

The invention provides a wind turbine generator fault early warning method and system based on a sudden change characterization value. The method comprises the steps that vibration data and operation parameters of a wind turbine generator are acquired; storing the vibration data into a time sequence array with a fixed length, updating the array by replacing the data, and calculating the standard deviation and kurtosis of the vibration data in the array in real time; extracting the characteristics of the operation parameters, and identifying the operation state of the unit; when the wind turbine generator set is in a normal operation state, weighted summation is carried out based on the standard deviation and kurtosis of the vibration data, and a vibration data mutation characterization value is obtained; and when the vibration data mutation characterization value and the kurtosis of the vibration data are both larger than the corresponding preset safety threshold values, the wind turbine generator is controlled to stop, and an alarm signal is uploaded. According to the method, the vibration sudden change characterization value is introduced, safety protection control is started in time, and major secondary disasters such as tower sweeping and tower falling caused after the unit breaks down are effectively prevented.

Description

Technical Field

[0001] The present invention relates to the technical field of wind turbine fault warning, and in particular to a wind turbine fault warning method and system based on mutation characterization values. Background Art

[0002] Wind turbines are inevitably subject to abnormal damage such as tower corrosion, material degradation, bolt loosening, and blade failure due to long-term operation in complex environments. These problems may have a negative impact on the performance and reliability of wind turbines. In severe cases, major safety accidents such as tower collapse may occur, which can cause tens of millions of economic losses and even casualties. In recent years, wind turbine tower collapse accidents have occurred frequently, so it is of great significance to carry out safety protection and control of wind turbines to prevent tower collapse.

[0003] Most of the existing vibration analysis technical methods use vibration data amplitude judgment and vibration frequency change trend identification: when a sudden increase in the vibration data amplitude is detected, the unit is protected against tower collapse. However, when the vibration data threshold suddenly increases, the wind turbine has already suffered a serious accident, and the protection at this time cannot achieve the purpose of safety protection; and the trend change of vibration frequency can only monitor the operating status of the unit and determine whether the unit is running with a problem, but cannot directly characterize the risk of tower collapse, and cannot effectively identify sudden unit serious accidents. Summary of the invention

[0004] In order to solve the above problems, the present invention proposes a wind turbine fault warning method and system based on mutation characterization values. By introducing mutation characterization values ​​of vibration data, the risk of tower collapse and tower sweeping can be characterized, and the degree of change of vibration data of the wind turbine can be effectively identified. Timely and effective warning can be given before the accident occurs, thereby preventing the unit from secondary disasters caused by serious accidents and preventing tower collapse accidents of wind turbines.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] In a first aspect, the present invention provides a wind turbine fault early warning method based on a mutation characterization value, comprising:

[0007] Obtain wind turbine vibration data and operating parameters;

[0008] Storing the vibration data in a fixed-length time series array, updating the array by replacing the data, and calculating the standard deviation and kurtosis of the vibration data in the array in real time;

[0009] Extracting the characteristics of the operating parameters and identifying the operating status of the unit;

[0010] When the wind turbine is in normal operation, a weighted sum is performed based on the standard deviation and kurtosis of the vibration data to obtain a vibration data mutation characterization value;

[0011] When the mutation characterization value of the vibration data and the kurtosis of the vibration data are both greater than the corresponding preset safety thresholds, the wind turbine is controlled to shut down and an alarm signal is uploaded.

[0012] Preferably, the vibration data includes blade vibration data and cabin vibration data; the operating parameters include wind speed, power, rotation speed, cabin position, pitch position and unit status position data.

[0013] Preferably, the vibration data is stored in a time series array of fixed length, and the array is updated by replacing the data, specifically:

[0014] Constructing an initial array, storing the vibration data collected in the current period into the initial array; the array is a time series of fixed length;

[0015] In each new cycle, one vibration data in the array is updated from left to right.

[0016] Preferably, the standard deviation and kurtosis of the vibration data in the array are calculated in real time, specifically:

[0017]

[0018] Among them, S n is the standard deviation, K n is the kurtosis, a i is the vibration data, is the average of the vibration data array, and n represents the number of data.

[0019] Preferably, the extracting the features of the operating parameters and identifying the operating status of the unit is specifically:

[0020] According to the cut-in wind speed and cut-out wind speed thresholds of the wind turbine, real-time monitoring is performed to determine whether the current wind speed of the wind turbine is within the operating wind speed;

[0021] Real-time monitoring of wind turbine power values ​​and unit status, identifying whether the wind turbine is currently operating in a power generation state;

[0022] Monitor the position of the wind turbine nacelle in real time and identify whether the wind turbine is currently in a yaw state.

[0023] Preferably, when the wind turbine is in a normal operating state, a vibration data mutation characterization value is obtained by performing weighted summation based on the standard deviation and kurtosis of the vibration data, specifically:

[0024] η=μS n +λK n

[0025] Among them, μ is the weighting coefficient of the standard deviation of vibration data, and λ is the weighting coefficient of the kurtosis of vibration data.

[0026] Preferably, when the vibration data mutation characterization value and the kurtosis of the vibration data are both greater than the corresponding preset safety thresholds, the wind turbine is controlled to shut down and an alarm signal is uploaded, specifically:

[0027] When the vibration data mutation characterization value is greater than the preset safety threshold, a separate threshold judgment is performed on the vibration data kurtosis;

[0028] If the vibration data mutation characterization value and the vibration data kurtosis both exceed the corresponding preset safety thresholds, the wind turbine is controlled to enter rapid shutdown and an alarm signal is uploaded for safety protection control.

[0029] In a second aspect, the present invention provides a wind turbine fault warning system based on mutation characterization values, comprising:

[0030] A data acquisition module, used to acquire vibration data and operating parameters of the wind turbine generator set;

[0031] A difference calculation module, used for storing the vibration data into a time series array of fixed length, updating the array by replacing the data, and calculating the standard deviation and kurtosis of the vibration data in the array in real time;

[0032] A state recognition module, used to extract the characteristics of the operating parameters and identify the unit operating state;

[0033] A characterization calculation module is used to obtain a vibration data mutation characterization value by weighted summation based on the standard deviation and kurtosis of the vibration data when the wind turbine is in normal operation;

[0034] The early warning module is used to control the wind turbine to shut down and upload an alarm signal when the vibration data mutation characterization value and the kurtosis of the vibration data are both greater than the corresponding preset safety thresholds.

[0035] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a wind turbine fault warning method based on mutation characterization values ​​described in the first aspect.

[0036] In a fourth aspect, the present invention provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of a wind turbine fault warning method based on mutation characterization values ​​described in the first aspect are implemented.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] (1) The present invention constructs a comprehensive monitoring foundation by acquiring the vibration data and operating parameters of the wind turbine generator set. The vibration data is stored in a fixed-length time series array and the standard deviation and kurtosis are calculated in real time, which can dynamically capture the characteristics of the vibration data. The operating parameter characteristics are extracted to identify the operating status of the unit, making the fault judgment more accurate. When the unit is operating normally, the mutation characterization value is obtained based on the weighted sum of the standard deviation and kurtosis. This method is more scientific than the traditional method of judging only based on the amplitude of the vibration data. When the vibration data threshold suddenly increases in the traditional method, the unit often has a serious accident. The present invention can provide early warning by monitoring the mutation characterization value in real time; and when the mutation characterization value is greater than the preset safety threshold, the kurtosis is also judged separately. If both exceed the threshold, the rapid shutdown is controlled, which effectively avoids the occurrence of serious accidents, overcomes the defect that the existing technology cannot effectively identify sudden serious accidents of the unit, and ensures the safe operation of the wind turbine.

[0039] (2) The wind turbine tower collapse prevention control method based on vibration data statistical indicators provided by the present invention does not rely on SCADA to obtain data and can interact with PLC control. Through real-time monitoring of vibration kurtosis and vibration standard deviation, the sudden change state of unit vibration can be identified and tower collapse prevention protection control can be performed. At the same time, the present invention realizes online identification of sudden change state of unit vibration and has real-time control.

[0040] (3) The wind turbine tower collapse prevention control method based on vibration data statistical indicators provided by the present invention receives wind speed, power, rotation speed, cabin position, pitch position, and unit status data, and archives and stores the data obtained at the corresponding time point to provide data support for subsequent detection and analysis, which is helpful to better carry out tower collapse prevention control related work.

[0041] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their description are used to explain the present invention but do not constitute a limitation of the present invention.

[0043] Figure 1 A main flow chart of a wind turbine fault early warning method based on mutation characterization values ​​provided by an embodiment of the present invention;

[0044] Figure 2 A schematic diagram of vibration mutation characterization values ​​provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0046] In the field of wind turbine safety, tower collapse and tower sweeping are unique and urgent compared to other conventional faults. Some electrical or control system faults may show obvious abnormal symptoms before the fault occurs, such as abnormal sensor readings, system alarms, etc., making them easier to detect and prevent early. In contrast, tower collapse and tower sweeping accidents are often more sudden. Since these faults are usually related to long-term accumulated structural damage, material aging or extreme weather conditions, there are often no obvious warning signals or only weak vibration frequency changes before they occur, which makes it difficult for traditional monitoring methods based on vibration data amplitude judgment and frequency change trends to effectively predict.

[0047] In order to solve the above problems, the present invention calculates a mutation characterization value to perform timely and effective accident prediction before tower collapse and tower sweeping accidents occur.

[0048] Embodiment 1

[0049] like Figure 1 As shown, this embodiment discloses a wind turbine fault warning method based on mutation characterization value, comprising the following steps:

[0050] S1: Obtain vibration data and operating parameters of wind turbine generator set;

[0051] S2: storing the vibration data into a time series array of fixed length, updating the array by replacing the data, and calculating the standard deviation and kurtosis of the vibration data in the array in real time;

[0052] S3: extracting the characteristics of the operating parameters and identifying the operating status of the unit;

[0053] S4: When the wind turbine is in normal operation, a weighted sum is performed based on the standard deviation and kurtosis of the vibration data to obtain a vibration data mutation characterization value;

[0054] S5: When the vibration data mutation characterization value and the kurtosis of the vibration data are both greater than the corresponding preset safety thresholds, the wind turbine is controlled to shut down and an alarm signal is uploaded.

[0055] Next, combine Figure 1 , a wind turbine fault warning method based on mutation characterization value disclosed in this embodiment is described in detail.

[0056] In S1, the vibration data includes blade vibration data and cabin vibration data; the operating parameters include wind speed, power, rotation speed, cabin position, pitch position and unit status position data.

[0057] S2 specifically includes:

[0058] S201: constructing an initial array, storing the vibration data collected in the current period into the initial array; the array is a time series of fixed length;

[0059] S202: In each new cycle, a vibration data in the array is updated cyclically from left to right.

[0060] It should be further explained that in step 2, an empty array A with a fixed length of n is first created. n When the main control system is executed, the vibration data a is stored in each cycle i , until the array A n Fill up; then update the vibration data array in real time in each execution cycle, the update method is as follows:

[0061] A n =[a1,a2,a3,...,a n ];

[0062] A n+1 =[a n+1 ,a2,a3,...,a n ];

[0063] A n+2 =[a n+1 ,a n+2 ,a3,...,a n ];

[0064] .......

[0065] The vibration data is stored in an array of fixed length, and the standard deviation and kurtosis of the vibration data in the array are calculated in real time by replacing them one by one.

[0066] By continuously replacing old data, the array of this embodiment can always contain the latest vibration information, and the array will not contain repeated or redundant information, which is conducive to ensuring the accuracy of data analysis and reducing the computational overhead caused by redundant data. In real-time systems, the latest data is usually the most valuable. By replacing the data to update the array, it can be ensured that the system always makes decisions based on the latest information, thereby improving the response speed and accuracy of the system. At the same time, the fixed-length array limits the occupancy of storage space. As new data arrives, the old data is replaced, thereby avoiding unlimited growth in the amount of data. In addition, since the array length is fixed, the algorithm does not need to consider the problem of dynamically adjusting the array size when processing data, thereby simplifying the processing logic and improving efficiency, and the fixed-length array is also easier to index and access.

[0067] The main control system is based on the vibration data array A in real time n Calculate the standard deviation S of vibration data and the kurtosis K of vibration data;

[0068]

[0069] in, is the average of the vibration data array, a i is the vibration data, and n represents the number of data; wherein the vibration data includes blade vibration data and cabin vibration data, which are respectively input into the above formula for calculation, and any type of vibration data can be used for fault judgment.

[0070] In S3, the characteristics of the operating parameters are extracted to identify the operating status of the unit, specifically:

[0071] S301: monitoring in real time whether the current wind speed of the wind turbine is at the operating wind speed according to the cut-in wind speed and cut-out wind speed thresholds of the wind turbine;

[0072] S302: Real-time monitoring of the wind turbine power value and the status of the wind turbine to identify whether the wind turbine is currently in a power generation state;

[0073] S303: Monitor the position of the wind turbine nacelle in real time to identify whether the current operating state of the wind turbine is in a yaw state.

[0074] Preferably, the normal operating state of the wind turbine is:

[0075] (1) Cut-in wind speed <W s <Cut out wind speed;

[0076] (2)P>10kw;

[0077] (3)N P =Nplast;

[0078] (4)P P <Ppset;

[0079] (5)S t = Power generation status;

[0080] Among them, W s is the current measured wind speed, P is the unit power, N P is the cabin position, Nplast is the cabin position of the previous cycle, P P is the pitch position, Ppset is the pitch setting threshold for entering the protection state, S t It is the unit status bit.

[0081] The parameters selected in this embodiment can fully cover the operating characteristics of the wind turbine, including its external environment (such as wind speed), internal working status (such as power, speed), physical position (such as cabin position, pitch position) and overall operating status (such as unit status position). By real-time monitoring of these parameters, a comprehensive understanding and grasp of the operating status of the wind turbine can be achieved, thereby improving the accuracy of fault warning.

[0082] In S4, when the wind turbine is in normal operation, a vibration data mutation characterization value is obtained by weighted summation based on the standard deviation and kurtosis of the vibration data. The vibration data mutation characterization value is specifically:

[0083] η=μS n +λK n

[0084] Among them, μ is the weighting coefficient of the standard deviation of vibration data, and λ is the weighting coefficient of the kurtosis of vibration data.

[0085] It is further necessary to explain that if Figure 2 As shown in the figure, vibration kurtosis is easier to characterize the sudden change of vibration data, but using vibration kurtosis alone for threshold judgment will cause the unit to falsely trigger the alarm. The use of vibration kurtosis and vibration standard deviation weighted threshold judgment can more accurately identify the sudden change state of vibration of the unit. The order of magnitude of vibration kurtosis and vibration standard deviation is different. Under normal circumstances, the value of vibration kurtosis is about 3, and the value of vibration standard deviation is about 0.12. The weighted coefficient μ of vibration data standard deviation and the weighted coefficient λ of vibration data kurtosis are set so that the weighted proportion of vibration kurtosis is 70% and the weighted proportion of vibration standard deviation is 30%. Finally, the vibration sudden change characterization value is output.

[0086] In S5, a threshold judgment is performed on the vibration mutation characterization value. When the vibration mutation characterization value and the vibration data kurtosis are both greater than the corresponding preset safety thresholds, the main control system controls the wind turbine to shut down quickly and uploads the unit vibration mutation fault signal to the HMI system for real-time display.

[0087] Specifically, a threshold is determined for the vibration mutation characterization value. When the vibration mutation characterization value is greater than the preset safety threshold and the vibration data kurtosis also exceeds its preset value, the main control system controls the wind turbine to shut down quickly, uploads the unit vibration mutation fault signal to the HMI system for alarm and real-time display, and remotely resets through the HMI system. It should be understood that the preset safety threshold is set by those skilled in the art according to actual conditions.

[0088] Kurtosis is more sensitive to abnormal mutations in data, and can better capture those tiny changes in the signal that may contain fault information than standard deviation. It can be seen that the high sensitivity of kurtosis enables it to quickly capture potential faults, but it is also prone to false alarms. In order to balance the sensitivity of kurtosis and reduce false alarms of unit faults, this embodiment adds a characterization value on the basis of considering kurtosis. Characterization values ​​and kurtosis reflect data characteristics from different angles. Characterization values ​​integrate multiple information of the data, while kurtosis highlights abnormal mutations. The combination of the two can more comprehensively evaluate the data status, complement and verify each other, and improve the accuracy and reliability of fault judgment.

[0089] This embodiment aims to solve the problem of inaccurate identification of sudden faults of wind turbines and low safety in the prior art. This specific embodiment introduces a vibration data mutation characterization value and adopts vibration kurtosis and vibration standard deviation weighted threshold judgment to more accurately identify the sudden vibration state of the unit, provide early warning before a fault occurs, and improve the wind turbine's tower prevention and tower cleaning capabilities.

[0090] Embodiment 2

[0091] This embodiment provides a wind turbine fault warning system based on mutation characterization values, including:

[0092] A data acquisition module, used to acquire vibration data and operating parameters of the wind turbine generator set;

[0093] A difference calculation module, used for storing the vibration data into a time series array of fixed length, updating the array by replacing the data, and calculating the standard deviation and kurtosis of the vibration data in the array in real time;

[0094] A state recognition module, used to extract the characteristics of the operating parameters and identify the unit operating state;

[0095] A characterization calculation module is used to obtain a vibration data mutation characterization value by weighted summation based on the standard deviation and kurtosis of the vibration data when the wind turbine is in normal operation;

[0096] The early warning module is used to control the wind turbine to shut down and upload an alarm signal when the vibration data mutation characterization value and the kurtosis of the vibration data are both greater than the corresponding preset safety thresholds.

[0097] Embodiment 3

[0098] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps of the wind turbine fault early warning method based on mutation characterization values ​​as described in the first embodiment above are implemented.

[0099] Embodiment 4

[0100] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps of a wind turbine fault warning method based on mutation characterization values ​​as described in the first embodiment are implemented.

[0101] The steps or modules involved in the above embodiments 2 to 4 correspond to those in embodiment 1. For the specific implementation, please refer to the relevant description of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.

[0102] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A wind turbine fault early warning method based on mutation characterization value, characterized in that: include: Obtain wind turbine vibration data and operating parameters; Storing the vibration data in a fixed-length time series array, updating the array by replacing the data, and calculating the standard deviation and kurtosis of the vibration data in the array in real time; Extracting the characteristics of the operating parameters and identifying the operating status of the unit; When the wind turbine is in normal operation, a weighted sum is performed based on the standard deviation and kurtosis of the vibration data to obtain a vibration data mutation characterization value; When the mutation characterization value of the vibration data and the kurtosis of the vibration data are both greater than the corresponding preset safety thresholds, the wind turbine is controlled to shut down and an alarm signal is uploaded.

2. A wind turbine fault early warning method based on mutation characterization value according to claim 1, characterized in that: The vibration data includes blade vibration data and cabin vibration data; the operating parameters include wind speed, power, rotation speed, cabin position, pitch position and unit status position data.

3. A wind turbine fault early warning method based on mutation characterization value according to claim 1, characterized in that: The vibration data is stored in a time series array of fixed length, and the array is updated by replacing the data, specifically: Constructing an initial array, storing the vibration data collected in the current period into the initial array; the array is a time series of fixed length; In each new cycle, one vibration data in the array is updated from left to right.

4. A wind turbine fault early warning method based on mutation characterization value according to claim 1, characterized in that: Calculate the standard deviation and kurtosis of the vibration data in the array in real time, specifically: Among them, S n is the standard deviation, K n is the kurtosis, a i is the vibration data, is the average of the vibration data array, and n represents the number of data.

5. A wind turbine fault early warning method based on mutation characterization value as claimed in claim 2, characterized in that: The extracting of the characteristics of the operating parameters and identifying the operating status of the unit is specifically as follows: According to the cut-in wind speed and cut-out wind speed thresholds of the wind turbine, real-time monitoring is performed to determine whether the current wind speed of the wind turbine is within the operating wind speed; Real-time monitoring of wind turbine power values ​​and unit status, identifying whether the wind turbine is currently operating in a power generation state; Monitor the position of the wind turbine nacelle in real time and identify whether the wind turbine is currently in a yaw state.

6. A wind turbine fault early warning method based on mutation characterization value according to claim 1, characterized in that: When the wind turbine is in normal operation, a vibration data mutation characterization value is obtained by weighted summing based on the standard deviation and kurtosis of the vibration data, which is specifically: η=μS n +λK n Among them, μ is the weighting coefficient of the standard deviation of vibration data, and λ is the weighting coefficient of the kurtosis of vibration data.

7. A wind turbine fault early warning method based on mutation characterization value according to claim 1, characterized in that: When the vibration data mutation characterization value and the kurtosis of the vibration data are both greater than the corresponding preset safety threshold, the wind turbine is controlled to shut down and an alarm signal is uploaded, specifically: When the vibration data mutation characterization value is greater than the preset safety threshold, a separate threshold judgment is performed on the vibration data kurtosis; If the vibration data mutation characterization value and the vibration data kurtosis both exceed the corresponding preset safety thresholds, the wind turbine is controlled to enter rapid shutdown and an alarm signal is uploaded for safety protection control.

8. A wind turbine fault warning system based on mutation characterization value, characterized in that: include: A data acquisition module, used to acquire vibration data and operating parameters of the wind turbine generator set; A difference calculation module, used for storing the vibration data into a time series array of fixed length, updating the array by replacing the data, and calculating the standard deviation and kurtosis of the vibration data in the array in real time; A state recognition module, used to extract the characteristics of the operating parameters and identify the unit operating state; A characterization calculation module is used to obtain a vibration data mutation characterization value by weighted summation based on the standard deviation and kurtosis of the vibration data when the wind turbine is in normal operation; The early warning module is used to control the wind turbine to shut down and upload an alarm signal when the vibration data mutation characterization value and the kurtosis of the vibration data are both greater than the corresponding preset safety thresholds.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by the processor, the steps of the wind turbine fault early warning method based on mutation characterization values ​​as described in any one of claims 1 to 7 are implemented.

10. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the wind turbine fault warning method based on mutation characterization value as described in any one of claims 1-7 are implemented.

Images