A wind turbine generator bearing fault early warning method and device

Abstract

The present invention relates to a method and device for early warning of a generator bearing fault of a wind turbine generator, comprising the steps of: screening historical data of the generator bearing of the wind turbine generator; obtaining a scatter plot of the relationship between the temperature rise and the rotational speed of the generator bearing; obtaining a curve of the relationship between the temperature rise and the rotational speed of the generator bearing; calculating the area enclosed by the relationship curve and the initial temperature horizontal line; constructing a detection index K; calculating past indicators; and evaluating the health status. The beneficial effects of the present invention are: using historical data, introducing ambient temperature and generator rotational speed, and establishing a new feature of the temperature difference between the front and rear end bearings that can characterize the independent characteristics of each unit generator bearing; by monitoring the temperature rise rate and the temperature difference between the front and rear end bearings, the sensitivity of the fault early warning device to the early failure of the bearing is effectively improved; a fault monitoring index for the generator bearing is proposed, and the three monitoring indicators of monthly, biweekly, and weekly are used to comprehensively evaluate the health status of the bearing, effectively improving the accuracy of the bearing fault early warning.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind power, and in particular relates to a method and a device for early warning of a bearing fault of a wind turbine generator. Background Art

[0002] Large-scale and high-quality development of the wind power industry is an important strategic choice for implementing the dual carbon goals and tasks, but the high cost of operation and maintenance is an important factor that currently hinders the accelerated and healthy development of the wind power industry. Industry surveys and statistics show that the failure of the generator bearing of a wind turbine causes a long period of unplanned downtime of the unit, and accounts for a high proportion of the maintenance cost of the unit. In-depth investigations and studies have found that most wind farms currently use scheduled maintenance and post-fault maintenance methods, which results in potential faults not being discovered in time, and can only passively carry out unplanned maintenance and replacement after the fault shutdown; this not only increases the cost required for maintenance and replacement, but also causes a loss of power generation. In addition, repeated maintenance of intact equipment will also increase operation and maintenance costs, and even cause excessive maintenance, affecting the service life of the equipment and greatly increasing the operation and maintenance costs. Therefore, it is of great significance to study and develop a method and device for early warning of generator bearing faults of wind turbines.

[0003] At present, the industry mainly achieves the purpose of alarm by setting bearing temperature limits in the wind turbine main control system. However, due to the large number of wind turbine models and the differences in structural loads of different units, it is often impossible to detect bearing failures in a timely and accurate manner by setting alarm limits. Moreover, even for units of the same model in the same wind farm, there will be individual differences in bearing temperature changes due to differences in wind load, environment, and maintenance conditions.

[0004] In summary, the current wind turbine generator bearing fault warning methods have the following problems: 1) The bearing temperature alarm limit is not universal, which can easily lead to missed faults and false alarms; 2) The simple temperature threshold comparison method has low sensitivity to early bearing failures; 3) The bearing temperature alarm limit is set loosely, resulting in many bearing failures not being discovered in time. Summary of the invention

[0005] The purpose of the present invention is to overcome the deficiencies in the prior art and to provide a method and device for early warning of a bearing fault of a wind turbine generator.

[0006] The wind turbine generator bearing fault warning device includes an operation data acquisition unit for collecting data, a data storage unit for storing data, a data processing and analysis unit for processing and analyzing data, a threshold monitoring and alarm unit for determining the health status of the generator bearing, and an operation and maintenance feedback unit for prompting and feeding back maintenance information;

[0007] The operation data acquisition unit, data storage unit, data processing and analysis unit, threshold monitoring and alarm unit and operation and maintenance feedback unit are electrically connected in sequence; the data processing and analysis unit and the threshold monitoring and alarm unit are both electrically connected to the data storage unit, and the operation and maintenance feedback unit is electrically connected to the threshold monitoring and alarm unit.

[0008] Preferably, in each electrically connected line, the signal flows in one direction.

[0009] The early warning method of the wind turbine generator bearing fault early warning device comprises the following steps:

[0010] Step 1, obtaining historical data of the generator drive end bearing temperature, the generator non-drive end bearing temperature, the fan generator speed, the fan cabin temperature and the fan status, and eliminating abnormal data and invalid data;

[0011] Step 2, obtaining the generator bearing temperature rise data and drawing a scatter plot of the relationship between the generator bearing temperature rise and the rotation speed;

[0012] Step 3, respectively obtain the relationship curve L1 between the bearing temperature rise at the driving end of the generator and the speed, the relationship curve L2 between the bearing temperature rise at the non-driving end of the generator and the speed, and the relationship curve L3 between the temperature difference between the driving end and the non-driving end of the generator and the speed;

[0013] Step 4: Calculate the area S enclosed by L1, L2 and L3 and the corresponding initial temperature horizontal line d , S Nd , S e ;

[0014] Step 5: Construct the generator bearing health monitoring indicator K:

[0015] K=λ1S e +λ2S d +λ3S Nd

[0016] Among them, λ1,λ2,λ3∈[0,1] are temperature index adjustment parameters;

[0017] Step 6: Calculate the operating data for one month, two weeks, and one week before the current time to obtain the monthly monitoring value K M , biweekly monitoring value K 2W and the weekly monitoring value K W ; and send the results to the data storage unit and the threshold monitoring alarm unit;

[0018] Step 7. Compare K M , K 2W , K W and threshold Limit_K M 、Limit_K 2W、Limit_K w ;

[0019] Step 8: Evaluate the health status of the wind turbine generator bearing and send it to the data storage unit.

[0020] As a preferred embodiment, in step 1: filter out historical data when the unit is operating normally and generating electricity without faults according to the status of the fan; set the upper and lower limits of the speed according to the characteristics of the generator speed range, and the data volume within the selected speed range accounts for greater than or equal to 90%; finally, eliminate abnormal points in the data at twice the preset upper temperature limit of the SCADA system.

[0021] Preferably, step 2 is specifically as follows:

[0022] Generator drive end bearing temperature rise = generator drive end bearing temperature - fan cabin temperature;

[0023] Generator non-drive end bearing temperature rise = generator non-drive end bearing temperature - fan cabin temperature;

[0024] The temperature difference between the generator drive end and the non-drive end = the generator drive end bearing temperature - the generator non-drive end bearing temperature;

[0025] According to the calculation results of the above formula, the scatter diagram of the relationship between the bearing temperature rise at the driving end of the generator and the speed, the scatter diagram of the relationship between the bearing temperature rise at the non-driving end of the generator and the speed, and the scatter diagram of the relationship between the temperature difference between the driving end and the non-driving end of the generator and the speed are obtained respectively.

[0026] Preferably, step 3 is specifically as follows: segmenting the rotational speed into multiple intervals, calculating the average value of the generator bearing temperature rise data points in each interval, connecting the average values ​​of each speed segment into a curve, and obtaining a curve of the relationship between the generator bearing temperature rise and the rotational speed.

[0027] Preferably, in step 4: the height of the initial temperature horizontal line is the temperature rise or temperature difference of the generator bearing at the lowest speed.

[0028] As a preferred method, in step 6: the monthly monitoring value K M , biweekly monitoring value K 2W and the weekly monitoring value K W The calculation formulas are:

[0029]

[0030] in The areas enclosed by the curves L3, L1, L2 formed by the data from the current moment to the previous month and their initial temperature horizontal lines; The areas enclosed by the curves L3, L1, L2 formed by the data two weeks before the current moment and their initial temperature horizontal lines; They are the areas enclosed by the curves L3, L1, L2 formed by the data one week before the current moment and their initial temperature horizontal lines.

[0031] Preferably, in step 7: K M , K 2W and K W and threshold Limit_K M 、Limit_K 2W 、Limit_K w The comparison is performed in the threshold monitoring alarm unit 14, where Limit_K M 、Limit_K 2W 、Limit_K w All of them are custom parameters, which are optimized based on the early warning feedback results.

[0032] As a preferred method, in step 8: the following evaluation formula is used to evaluate the health status of the generator bearing of the wind turbine generator:

[0033]

[0034] When the evaluation result shows attention, warning or alarm, the threshold monitoring alarm unit sends a command signal to the operation and maintenance feedback unit; after receiving the command, the operation and maintenance feedback unit inspects the corresponding wind turbine generator bearing and feeds back the result to the threshold monitoring alarm unit, and the command signal of the threshold monitoring alarm unit stops.

[0035] The beneficial effects of the present invention are:

[0036] 1) The calculation model used in the present invention introduces two key factors that affect the bearing temperature: ambient temperature and generator speed. Different from the single fixed hard threshold method used in traditional generator bearing fault warning, it uses historical data to establish new features related to the temperature rise rate, and establishes a new feature of the front and rear end bearing temperature difference that can characterize the independent characteristics of each unit generator bearing.

[0037] 2) The present invention effectively improves the sensitivity of the fault warning device to early bearing failures by monitoring the temperature rise rate and the temperature difference between the front and rear end bearings, thereby avoiding the occurrence of serious failures.

[0038] 3) The present invention proposes fault monitoring indicators for generator bearings, and comprehensively utilizes three monitoring indicators, namely monthly, biweekly and weekly, to conduct a comprehensive assessment of the health status of the bearings, effectively improving the accuracy of bearing fault warning. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A detailed flow chart of a method for early warning of a generator bearing failure in a wind turbine;

[0040] Figure 2 It is a schematic diagram of the temperature rise curve of the driving end bearing of the wind turbine generator;

[0041] Figure 3 It is a schematic diagram of the area enclosed by the temperature rise curve L1 of the driving end bearing of the wind turbine generator;

[0042] Figure 4 A schematic diagram of the area enclosed by the temperature rise curve L2 of the non-drive end bearing of the wind turbine generator;

[0043] Figure 5 It is a schematic diagram of the area enclosed by the curve L3 of the relationship between the temperature difference and the rotation speed between the driving end and the non-driving end of the wind turbine generator;

[0044] Figure 6 The schematic diagram of the structure of the bearing fault early warning device of the wind turbine generator is shown in FIG.

[0045] Explanation of the reference numerals: operation data acquisition unit 11 , data storage unit 12 , data processing and analysis unit 13 , threshold monitoring and alarm unit 14 , operation and maintenance feedback unit 15 . DETAILED DESCRIPTION

[0046] The present invention is further described below in conjunction with embodiments. The description of the following embodiments is only used to help understand the present invention. It should be noted that for ordinary persons in the art, without departing from the principle of the present invention, the present invention can also be modified in some ways, and these improvements and modifications also fall within the scope of protection of the claims of the present invention.

[0047] As an embodiment, a wind turbine generator bearing fault warning device based on historical operation data analysis obtains historical operation data such as the wind turbine generator drive end bearing temperature, the generator non-drive end bearing temperature, the wind turbine generator speed, and the temperature inside the wind turbine cabin through continuous updating and iteration for analysis. In order to eliminate the influence of ambient temperature, three new features are specially established: the generator drive end bearing temperature rise, the generator non-drive end bearing temperature rise, and the temperature difference between the front and rear ends of the generator bearing. On this basis, scatter plots of the above features changing with the generator speed are drawn respectively. The average value of each speed interval is obtained by taking the average speed segment by segment to establish the relationship curve of each feature with the speed, and then the area enclosed by each relationship curve and the initial temperature line is calculated. Finally, the above area is substituted into the monitoring index calculation formula to calculate the monthly monitoring value K respectively. M , biweekly monitoring value K 2W , Single-week monitoring value K W , each monitoring value is compared with its threshold value to evaluate the current health status of the generator bearing and output the result.

[0048] Figure 6The composition of the wind turbine generator bearing fault early warning device is shown, including an operation data acquisition unit 11, a data storage unit 12, a data processing and analysis unit 13, a threshold monitoring and alarm unit 14, and an operation and maintenance feedback unit 15, wherein the operation data acquisition unit 11 is electrically connected to the data storage unit 12, the data storage unit 12 is electrically connected to the data processing and analysis unit 13, the data processing and analysis unit 13 is electrically connected to the data storage unit 12 and the threshold monitoring and alarm unit 14, the threshold monitoring and alarm unit 14 is electrically connected to the data storage unit 12 and the operation and maintenance feedback unit 15, and the operation and maintenance feedback unit 15 is electrically connected to the threshold monitoring and alarm unit 14.

[0049] The operation data acquisition unit 11 is used to continuously obtain the latest operation data from the unit SCADA system, and can be timed or triggered to execute according to the evaluation requirements; the data storage unit 12 is used to store the historical operation data of the wind turbine and the result data generated by the subsequent data analysis. In this embodiment, in order to save hardware costs, the data storage unit 12 dynamically updates and removes old data so that the database only retains one and a half months of historical operation data; the data processing and analysis unit 13 is used to process the operation data and calculate the monitoring indicators, and the results of the calculation and analysis are stored in the unit 12; the threshold monitoring and alarm unit 14 is used to analyze and compare the relationship between the monitoring indicators and their corresponding thresholds, and output the judgment results of the bearing health status. The threshold parameters of the unit can also be customized in this unit; the operation and maintenance feedback unit 15 is used to guide on-site operation and maintenance personnel to inspect the generator bearings and record feedback to optimize the threshold parameters.

[0050] Embodiment 2

[0051] As another embodiment, the early warning method of the wind turbine generator bearing fault early warning device based on historical operation data analysis described in embodiment 1 is described in detail in FIG. Figure 1 , the method is specifically:

[0052] Step 1: Continuously update and obtain historical operation data of wind turbines from the wind turbine SCADA system, so that the database always maintains the latest data volume of one and a half months. The collected operation data must include characteristic parameters such as the generator drive end bearing temperature, the generator non-drive end bearing temperature, the wind turbine generator speed, the wind turbine cabin temperature, and the wind turbine status;

[0053] Process the historical operation data of the wind turbine generator, remove abnormal and invalid data, and retain the data under normal operation of the wind turbine generator;

[0054] Step 1.1 Filter out the data when the unit is operating normally and generating electricity without any fault according to the status of the wind turbine;

[0055] Step 1.2 sets the upper and lower limits of the speed according to the characteristics of the generator speed range to ensure that the data volume in this range accounts for no less than 90%, thereby eliminating the speed points with a small data volume through this step. The embodiment of the present invention sets the speed screening condition at 950-1400RPM;

[0056] Step 1.3: Eliminate abnormal points in the data by twice the upper temperature limit preset by the SCADA system;

[0057] Step 2: To eliminate the influence of ambient temperature, three new features related to the bearing temperature of the wind turbine generator are established, and a scatter plot of the relationship between the new features and the speed is obtained;

[0058] Step 2.1 Establish a new feature of the temperature rise of the bearing at the drive end of the wind turbine generator:

[0059] Generator drive end bearing temperature rise = generator drive end bearing temperature - fan cabin temperature

[0060] Establish a new characteristic of the temperature rise of the non-drive end bearing of the wind turbine generator:

[0061] Generator non-drive end bearing temperature rise = generator non-drive end bearing temperature - fan cabin temperature

[0062] Establish a new feature of the temperature difference between the drive end and non-drive end of the wind turbine generator:

[0063] The temperature difference between the front and rear ends of the generator bearing = the bearing temperature at the drive end of the generator - the bearing temperature at the non-drive end of the generator

[0064] Step 2.2 constructs a scatter plot of the relationship between the temperature rise and speed of the wind turbine generator bearing;

[0065] Construct a scatter plot of the relationship between the bearing temperature rise and the speed at the drive end of the wind turbine generator, such as Figure 2 shown; reference Figure 2 Method, a scatter plot of the relationship between the temperature rise and speed of the non-drive end bearing of the wind turbine generator was constructed; Figure 2 Method, a scatter plot of the relationship between the temperature difference between the driving end and the non-driving end of the wind turbine generator and the speed is constructed;

[0066] Step 3: Establish the relationship curve between bearing temperature rise and speed based on the above scatter plots. In the embodiment of the present invention, the speed segmented averaging method is adopted to obtain the average temperature rise of each segment, and then the average values ​​of each speed segment point are connected into a line. Specifically, the speed is segmented at intervals of 50 RPM, and the average value of the temperature rise data points falling into each interval segment of [950, 1000), [1000, 1050), [1050, 1100), ..., [1350, 1400] is calculated, such as Figure 2 As shown by the middle curve;

[0067] Step 3.1 According to the above speed segment averaging method, establish the relationship curve L1 between the bearing temperature rise and speed at the drive end of the wind turbine generator, such as Figure 3 As shown;

[0068] Step 3.2: According to the above speed segment averaging method, establish the relationship curve L2 between the non-drive end bearing temperature rise and the speed of the wind turbine generator, such as Figure 4 As shown;

[0069] Step 3.3: According to the above speed segment averaging method, establish the relationship curve L3 between the temperature difference between the driving end and the non-driving end of the wind turbine generator and the speed, such as Figure 5 As shown;

[0070] Step 4: Calculate the area enclosed by the curve and its initial temperature horizontal line based on the above relationship curve. In this embodiment, the initial temperature horizontal line is defined as the temperature rise or temperature difference at a speed of 950 RPM. Since the bearing temperature is greatly affected by the speed, when pitting, flaking, pitting, wear and other faults occur inside the bearing, the frictional heat generation phenomenon will be more significant, and the temperature rise will be more obvious as the speed increases. Therefore, the area enclosed by the temperature rise curve can be used to more sensitively detect early faults inside the bearing;

[0071] Step 4.1 Calculate the area S enclosed by the curve L1 of the relationship between the temperature rise and speed of the wind turbine generator drive end bearing and its initial temperature horizontal line d ;

[0072] Step 4.2 Calculate the area S enclosed by the curve L2 of the relationship between the temperature rise and speed of the non-drive end bearing of the wind turbine generator and its initial temperature horizontal line Nd ;

[0073] Step 4.3 Calculate the area S enclosed by the curve L3 of the relationship between the temperature difference and the speed between the driving end and the non-driving end of the wind turbine generator and its initial temperature horizontal line. e ;

[0074] Step 5: Construct the generator bearing health monitoring indicator K:

[0075] K=λ1S e +λ2S d +λ3S Nd

[0076] In the above formula, λ1,λ2,λ3∈[0,1] are temperature index adjustment parameters, which can be continuously optimized and adjusted according to the early warning feedback. In this embodiment, λ1=0.5,λ2,λ3=0.25;

[0077] Step 6: To avoid the influence of occasional random factors, this embodiment comprehensively evaluates the health monitoring index K of the bearing from multiple time dimensions. Specifically, calculate the monthly monitoring value K obtained from the operating data of the previous month at the current time. M Calculate the biweekly monitoring value K obtained from the current time and the previous two weeks of operation data 2w Calculate the single-week monitoring value K obtained from the current time and the previous week's operation data W ;

[0078]

[0079] in The areas enclosed by the curves L3, L1, L2 formed by the data from the current moment to the previous month and their initial temperature horizontal lines; The areas enclosed by the curves L3, L1, L2 formed by the data two weeks before the current moment and their initial temperature horizontal lines; They are the areas enclosed by the curves L3, L1, L2 formed by the data one week before the current moment and their initial temperature horizontal lines.

[0080] Step 7: The above monitoring value K M , K 2W and K W , respectively with their thresholds Limit_K M 、Limit_K 2W and Limit_K w Compare and observe whether it exceeds the limit; the above thresholds are continuously optimized based on the early warning feedback results;

[0081] Step 8: Based on the following evaluation formula, the current health status of the wind turbine generator bearing can be obtained to determine whether it has a fault.

[0082]

[0083] The wind turbine generator bearings that need maintenance and repair are processed, and the processing results are fed back to the threshold monitoring alarm unit 14.

Claims

1. A warning method for a bearing fault warning device of a wind turbine generator, characterized in that: The wind turbine generator bearing fault early warning device comprises: an operation data acquisition unit (11) for acquiring data, a data storage unit (12) for storing data, a data processing and analysis unit (13) for processing and analyzing data, a threshold monitoring and alarm unit (14) for determining the health status of the generator bearing, and an operation and maintenance feedback unit (15) for prompting and feeding back maintenance information; the operation data acquisition unit (11), the data storage unit (12), the data processing and analysis unit (13), the threshold monitoring and alarm unit (14), and the operation and maintenance feedback unit (15) are electrically connected in sequence; the data processing and analysis unit (13) and the threshold monitoring and alarm unit (14) are both electrically connected to the data storage unit (12), and the operation and maintenance feedback unit (15) is electrically connected to the threshold monitoring and alarm unit (14); the early warning method of the wind turbine generator bearing fault early warning device comprises the following steps: Step 1, obtaining historical data of the generator drive end bearing temperature, the generator non-drive end bearing temperature, the fan generator speed, the fan cabin temperature and the fan status, and eliminating abnormal data and invalid data; Step 2, obtaining the generator bearing temperature rise data and drawing a scatter plot of the relationship between the generator bearing temperature rise and the rotation speed; Generator drive end bearing temperature rise = generator drive end bearing temperature - fan cabin temperature; Generator non-drive end bearing temperature rise = generator non-drive end bearing temperature - fan cabin temperature; The temperature difference between the generator drive end and the non-drive end = the generator drive end bearing temperature - the generator non-drive end bearing temperature; According to the calculation results of the above formula, the scatter diagram of the relationship between the bearing temperature rise at the drive end of the generator and the speed, the scatter diagram of the relationship between the bearing temperature rise at the non-drive end of the generator and the speed, and the scatter diagram of the relationship between the temperature difference between the drive end and the non-drive end of the generator and the speed are obtained respectively; Step 3, segment the speed into multiple intervals, calculate the average value of the generator bearing temperature rise data points in each interval, connect the average values ​​of each speed segment into a curve, and obtain the relationship curve between the generator bearing temperature rise and the speed; respectively obtain the relationship curve L1 between the generator drive end bearing temperature rise and the speed, the relationship curve L2 between the generator non-drive end bearing temperature rise and the speed, and the relationship curve L3 between the generator drive end and non-drive end temperature difference and the speed; Step 4: Calculate the area S enclosed by L1, L2 and L3 and the corresponding initial temperature horizontal line d , S Nd , S e ; Step 5: Construct the generator bearing health monitoring indicator K: K=λ1S e +λ2S d +λ3S Nd Among them, λ1,λ2,λ3∈[0,1] are temperature index adjustment parameters; Step 6: Calculate the operating data for one month, two weeks, and one week before the current time to obtain the monthly monitoring value K M , biweekly monitoring value K 2W and the weekly monitoring value K W ; and sending the result to the data storage unit (12) and the threshold monitoring alarm unit (14); Step 7. Compare K M , K 2W , K W and threshold Limit_K M 、Limit_K 2W 、Limit_K w ; Step 8: Evaluate the health status of the wind turbine generator bearing and send it to the data storage unit (12).

2. The early warning method of the wind turbine generator bearing fault early warning device according to claim 1 is characterized in that: In step 1: filter out historical data when the unit is operating normally and generating electricity without faults according to the status of the fan; set the upper and lower limits of the speed according to the characteristics of the generator speed range, and the data volume within the selected speed range accounts for greater than or equal to 90%; finally, eliminate abnormal points in the data according to twice the preset upper temperature limit of the SCADA system.

3. The early warning method of the wind turbine generator bearing fault early warning device according to claim 1 is characterized in that: In step 4: the height of the initial temperature horizontal line is the temperature rise or temperature difference of the generator bearing at the lowest speed.

4. The early warning method of the wind turbine generator bearing fault early warning device according to claim 1, characterized in that: Step 6: Monthly monitoring value K M , biweekly monitoring value K 2W and the weekly monitoring value K W The calculation formulas are: in The areas enclosed by the curves L3, L1, L2 formed by the data from the current moment to the previous month and their initial temperature horizontal lines; The areas enclosed by the curves L3, L1, L2 formed by the data two weeks before the current moment and their initial temperature horizontal lines; They are the areas enclosed by the curves L3, L1, L2 formed by the data one week before the current moment and their initial temperature horizontal lines.

5. The early warning method of the wind turbine generator bearing fault early warning device according to claim 1, characterized in that: Step 7: K M , K 2W and K W and threshold Limit_K M 、Limit_K 2w 、Limit_K w The comparison is performed in the threshold monitoring alarm unit (14), where Limit_K M 、Limit_K 2W 、Limit_K w All of them are custom parameters, which are optimized based on the early warning feedback results.

6. The early warning method of the wind turbine generator bearing fault early warning device according to claim 1, characterized in that: In step 8: Use the following evaluation formula to evaluate the health status of the wind turbine generator bearings When the evaluation result shows attention, warning or alarm, the threshold monitoring alarm unit (14) sends a command signal to the operation and maintenance feedback unit (15); after receiving the command, the operation and maintenance feedback unit (15) inspects the corresponding wind turbine generator bearing and feeds back the result to the threshold monitoring alarm unit (14), and the command signal of the threshold monitoring alarm unit (14) stops.

Images