Method and system for identifying states of mechanical pump and electric pump of gearbox of wind turbine generator

Abstract

The invention discloses a wind turbine generator gearbox mechanical pump and electric pump state identification method and system, and the method comprises the steps: S1, constructing a lubrication system operation model through extracting gearbox inlet oil pressure, lubricating oil temperature and generator rotation speed data; s2, decoupling operation states of a mechanical pump and an electric pump of the gearbox based on the lubrication system operation model; s3, extracting a shutdown segment, and identifying the state of the mechanical pump by adopting a Pearson correlation coefficient; and S4, extracting front and back two-time duration data fragments of the electric pump in a low-speed or high-speed gear starting period, and classifying and identifying the state of the electric pump by using a DBSCAN clustering algorithm. The method has the advantages of low cost, accurate identification and the like.

Description

Technical Field

[0001] The present invention mainly relates to the technical field of wind power, and particularly relates to a method and system for identifying the states of a mechanical pump and an electric pump of a wind turbine gearbox. Background Art

[0002] During the operation of the unit, when the coupling is damaged, the motor is damaged, the gear pump is cracked, etc., the mechanical pump and the electric pump of the gearbox cannot pump out lubricating oil normally, and cannot provide effective power for the oil circulation of the lubrication system. Ineffective lubrication may lead to premature wear or even failure of the components of the gearbox, affecting the normal operation of the unit and the life of the gearbox. Therefore, accurately detecting the states of the mechanical pump and the electric pump is crucial for ensuring the safe operation of the wind turbine. Summary of the Invention

[0003] Aiming at the technical problems existing in the prior art, the present invention provides a method and system for identifying the states of a mechanical pump and an electric pump of a wind turbine gearbox, which do not require additional sensors, have low cost, and are accurately identified.

[0004] To solve the above technical problems, the technical solution proposed by the present invention is as follows:

[0005] A method for identifying the states of a mechanical pump and an electric pump of a wind turbine gearbox includes the steps of:

[0006] S1. By extracting the data of the oil pressure at the inlet of the gearbox, the lubricating oil temperature, and the generator speed, a lubrication system operation model is constructed;

[0007] S2. Based on the lubrication system operation model, the operating states of the mechanical pump and the electric pump of the gearbox are decoupled;

[0008] S3. Extract the shutdown segment, and use the Pearson correlation coefficient to identify the state of the mechanical pump;

[0009] S4. Extract the data segments of the front and back two times the duration during the opening period of the electric pump in the low-speed or high-speed gear, and use the DBSCAN clustering algorithm to classify and identify the state of the electric pump.

[0010] Preferably, in step S1, the constructed lubrication system operation model is specifically:

[0011] P = P E ·[1 + ΔA(T)] + P M

[0012] where P is the oil pressure at the inlet of the gearbox, P E is the pressure provided by the electric pump at a constant temperature T 0 ΔA(T) is the change value of the reaction rate of the oil at temperature T obtained by using the Arrhenius equation, and P MThe pressure provided by the mechanical pump.

[0013] Preferably, the operating state of the mechanical pump is decoupled: when the state of the electric pump remains unchanged and the gearbox cooling fan is not turned on, the oil pressure provided by the electric pump is relatively stable at this time, and the lubricating oil temperature does not change significantly. At this time, the change in the inlet oil pressure is only affected by the mechanical pump, and the functional relationship between the inlet oil pressure and the mechanical pump can be solved. Specifically:

[0014] P M = k M n

[0015] where k M is the conversion coefficient and n is the generator speed.

[0016] Preferably, the operating state of the electric pump is decoupled: using the wind turbine speed, combined with the functional relationship between the mechanical pump and the inlet oil pressure, the influence of the mechanical pump on the inlet oil pressure is eliminated; according to the Arrhenius equation, the influence of the lubricating oil temperature on the viscosity is corrected, and then the influence of the lubricating oil temperature on the inlet oil pressure is corrected; at this time, the change in the oil pressure is only affected by the change in the state of the electric pump, and the oil pressure values that can be provided by each gear of the electric pump are solved. Specifically:

[0017]

[0018] where, ΔA(T) = A(T) - A(T 0 )

[0019] According to the Arrhenius equation:

[0020]

[0021] where, k E is the reaction rate constant, A is the frequency factor, E a is the activation energy, R is the gas constant, and T is the absolute temperature.

[0022] Preferably, the specific process of step S3 is as follows:

[0023] S31. Start;

[0024] S32. Extract the shutdown segment;

[0025] S33. Determine whether the duration exceeds the set value t and whether the range of the generator speed is greater than n0; if so, proceed to the next step;

[0026] S34. Determine whether the high / low-speed electric pump status word has changed; if so, proceed to the next step;

[0027] S35. Detect the Pearson correlation coefficient between the generator speed and the oil pressure at the inlet of the gearbox; if the Pearson correlation coefficient is less than the set threshold, it is determined that the mechanical pump fails; otherwise, it is effective.

[0028] Preferably, in step S32, the shutdown segment is: when the main control status word C0 - Cn is triggered, the generator speed drops rapidly, and the speed range difference > n0.

[0029] Preferably, the specific process of step S4 is as follows:

[0030] S41. Start

[0031] S42. Exclude the part of the oil pressure provided by the mechanical pump from the detection data;

[0032] S43. Use the Arrhenius formula to correct the oil pressure to 40°C;

[0033] S44. Extract two - fold data segments before and after the opening period of the low / high - speed gear of the electric pump;

[0034] S45. Check whether the running opening duration of the electric pump is greater than the set threshold t, and at the same time detect whether the average speed of the generator is greater than the set value n0; if so, go to the next step;

[0035] S46. Check whether the total duration of the data segment is less than the set threshold t1; if not, go to the next step;

[0036] S47. Calculate the average oil pressure increment during the opening period of the low - speed or high - speed gear of the electric pump;

[0037] S48. Based on the average oil pressure increment, use the DBSCAN clustering algorithm to classify all the data of the electric pump units and identify different types of operating states;

[0038] S49. According to the DBSCAN classification result, determine whether the current electric pump unit belongs to an outlier; if so, determine that the electric pump is abnormal; if not, determine that the electric pump is normal.

[0039] The present invention also discloses a computer program product, including a computer program, and the computer program executes the steps of the above - described method when being run by a processor.

[0040] The present invention further discloses a computer - readable storage medium, on which a computer program is stored, and the computer program executes the steps of the above - described method when being run by a processor.

[0041] The present invention also discloses a computer device, including a memory and a processor connected to each other, and a computer program is stored on the memory, and the computer program executes the steps of the above - described method when being run by a processor.

[0042] Compared with the prior art, the advantages of the present invention are as follows:

[0043] The present invention analyzes the operating mechanism of the lubrication system of a wind turbine gearbox and the operating principles of mechanical pumps and electric pumps, derives the model of the lubrication system, and separates the state characteristics of mechanical pumps and electric pumps therefrom. Based on the existing sensor parameters in the lubrication system: main control state, generator speed, oil pressure at the inlet of the gearbox, oil temperature at the inlet of the gearbox, etc., after extracting effective segments according to the state recognition methods of mechanical pumps and electric pumps, the Pearson correlation coefficient and the DBSCAN algorithm are respectively used to process the segments, and the states of mechanical pumps and electric pumps are judged according to the algorithm results.

[0044] The method of the present invention mainly involves the field of operation state monitoring and fault diagnosis of wind turbine units. When there is no dedicated sensor for monitoring the equipment, the data of other sensors in the system are used for state recognition, such as the state recognition method of the mechanical pump and electric pump of the gearbox without sensors.

[0045] The present invention does not require additional sensors, simplifying the structure of the monitoring system; traditional monitoring systems rely on a large number of sensors to collect equipment operation data, which not only increases the complexity and cost of the system, but also is limited by the lifespan and accuracy of sensors. To save costs and simplify the structure of the monitoring system, improve the reliability and practicality of the system, the present invention utilizes other monitoring variables in the gearbox lubrication system to realize the state recognition of mechanical pumps and electric pumps without additional sensors.

[0046] The present invention improves the accuracy and reliability of the state recognition of mechanical pumps and electric pumps. Since the mechanism of the present invention is clear and the principles and processes of the analysis means are clear, the state recognition of mechanical pumps and electric pumps can be carried out on a more scientific and rigorous basis. Through a detailed analysis of the operating mechanisms of mechanical pumps and electric pumps, the functional relationships between oil pressure and the rotational speed of mechanical pumps, and between oil temperature and the viscosity of lubricating oil are established, and the oil pressure is decomposed accordingly. The states of mechanical pumps and electric pumps are successfully judged using the Pearson correlation coefficient and the DBSCAN clustering algorithm, and abnormal units are accurately identified. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a diagram of the operating mechanism of the gearbox lubrication system in the present invention.

[0048] Figure 2 It is a schematic diagram of the working principle of the gear pump in the present invention.

[0049] Figure 3 It is a flowchart of the method for recognizing the state of the mechanical pump in the present invention in an embodiment.

[0050] Figure 4 It is a flowchart of the method for recognizing the state of the electric pump in the present invention in an embodiment.

[0051] Figure 5 This is a schematic diagram of the DBSCAN clustering algorithm in the present invention. Detailed implementation manners

[0052] The present invention will be further described below in conjunction with the accompanying drawings of the specification and specific embodiments.

[0053] The method for identifying the states of the mechanical pump and the electric pump of a wind turbine gearbox provided by an embodiment of the present invention includes the steps of:

[0054] S1. By extracting the data of the oil pressure at the inlet of the gearbox, the lubricating oil temperature, and the generator speed, a lubrication system operation model is constructed;

[0055] Analysis of the working principles of the mechanical pump and the electric pump: The mechanical pump of the gearbox usually uses a gear pump. A gear pump is a common positive displacement pump and is usually used to transport liquids. The working principle of a gear pump is that through the meshing movement of gears, the liquid is sucked in, compressed, and discharged, thereby realizing the transportation of the liquid. The working principle of a gear pump is relatively simple but very effective. The working principle of the gear pump will be described in detail below.

[0056] As Figure 2 shown, a gear pump usually has two basic structures. As Figure 2 shown in (a) therein, one consists of two meshing gears and a pump housing, where one gear is called the driving gear and the other is called the driven gear; as Figure 2 shown in (b) therein, the other consists of a single gear and a pump housing slightly larger than the gear. The working principles of the two structures are basically the same. There are an inlet and an outlet inside the pump housing. The lubricating oil enters the pump housing through the inlet and is then pushed by the gears. The rotation of the gears causes the volume between the teeth to decrease. After the lubricating oil is forced to increase in pressure, it finally exits from the outlet.

[0057] From the working process of the gear pump, it can be found that the factors affecting the performance of the gear pump are mainly: the power of the driving gear, the compression process of the pump cavity volume, and the sealing performance of the pump housing. If any of the three has a problem, it will lead to insufficient lubricating oil flow provided by the mechanical pump, resulting in a low oil pressure of the lubricating oil finally entering the gearbox.

[0058] An electric pump usually consists of a motor and a pump body, and generates fluid pressure through the drive of the motor. Except that the power source is different from that of the mechanical pump, the principle of transporting the liquid is the same as that of the mechanical pump, and will not be elaborated here.

[0059] As Figure 1 shown, the operating mechanism of the gearbox lubrication system is described as follows:

[0060] When the external environment changes (usually the change in wind speed), the gearbox changes accordingly following the operating conditions of the wind turbine. The specific change conditions are asFigure 1 As shown, according to the increase (as shown by the red path in Figure 1 ) or decrease (as shown by the green path in Figure 1 ) of the wind speed, the inlet oil pressure changes correspondingly.

[0061] From the above process, it can be seen that the factors affecting the inlet oil pressure of the gearbox are divided into the following three categories: the oil pressure provided by the mechanical pump, the oil pressure provided by the electric pump, and the lubricating oil temperature. Based on the above factors, a lubrication system operation model can be constructed:

[0062] P = P E ·[1 + ΔA(T)] + P M

[0063] where P is the inlet oil pressure of the gearbox, P E is the pressure provided by the electric pump at a constant temperature T 0 , ΔA(T) is the change value of the reaction rate of the oil at temperature T obtained by using the Arrhenius equation, and P M is the pressure provided by the mechanical pump.

[0064] S2. Decouple and analyze the operating states of the mechanical pump and the electric pump of the gearbox based on the lubrication system operation model;

[0065] Since there is coupling among the oil pressure provided by the mechanical pump, the oil pressure provided by the electric pump, and the lubricating oil temperature, it is difficult to directly distinguish them. To identify the states of each component, decoupling is required. The specific process is as follows:

[0066] Decoupling of the mechanical pump state: When the state of the electric pump remains unchanged and the gearbox cooling fan is not turned on, the oil pressure provided by the electric pump is relatively stable at this time, and the lubricating oil temperature does not change significantly. At this time, the change in the inlet oil pressure is only affected by the mechanical pump, and the functional relationship between the inlet oil pressure and the mechanical pump can be solved.

[0067] Decoupling of the electric pump state: Using the wind turbine speed and combining the functional relationship between the mechanical pump and the inlet oil pressure, the influence of the mechanical pump on the inlet oil pressure can be eliminated. According to the Arrhenius equation, the influence of the lubricating oil temperature on the viscosity is corrected, and then the influence of the lubricating oil temperature on the inlet oil pressure is corrected. At this time, the change in the oil pressure is only affected by the change in the state of the electric pump, and the oil pressure values that can be provided by each gear of the electric pump can be solved.

[0068] Specifically, for the process of the unit shutting down and the gearbox cooling fan not being turned on, the oil temperature remains basically stable. At this time, the oil pressure change caused by the oil temperature can be ignored. At this time, the oil pressure provided by the mechanical pump is related to the generator speed. That is:

[0069] P M = k M n

[0070] where k M is the conversion coefficient and n is the generator speed.

[0071] For the electric pump:

[0072]

[0073] where ΔA(T) = A(T) - A(T 0 ).

[0074] According to the Arrhenius equation:

[0075]

[0076] where k E is the reaction rate constant, A is the frequency factor, E a is the activation energy, R is the gas constant, and T is the absolute temperature.

[0077] S3. As Figure 3 shown, use the Pearson correlation coefficient analysis to identify the mechanical pump state. The specific steps are as follows:

[0078] S31. Start: The process starts, and the detection and judgment of the mechanical pump state begin;

[0079] S32. Extract the continuous process of the main control status words C0 - Cn during the shutdown process: The system extracts the sequence of main control status words (C0 - Cn) during the shutdown process from the operation data for subsequent analysis and judgment.

[0080] S33. Judgment condition 1: Whether the duration exceeds the set value t and whether the range of the generator speed is greater than n0;

[0081] Check whether the change of the main control status word lasts for more than the set value t, and at the same time detect whether the range of the generator speed is greater than the set value n0; if yes (the condition is met), go to the next step; if no (the condition is not met), the process ends directly, and the data segment does not meet the requirements and is discarded;

[0082] S34. Judgment condition 2: Whether the status word of the high / low-speed electric pump changes;

[0083] Check whether the status word of the high / low-speed electric pump changes to judge the change of the electric pump operation state; if yes (the status word changes), go to the next step; if no (the status word does not change), directly enter the judgment of condition 3;

[0084] S35. Judgment condition 3: Whether the oil pressure at the gearbox inlet decreases proportionally with the decrease of the generator speed;

[0085] Detect whether the oil pressure value at the inlet of the gearbox decreases proportionally with the decrease in the generator speed; if so (the oil pressure change is consistent with the speed decrease), it is determined that the mechanical pump is operating normally, and the process ends; if not (the oil pressure change is inconsistent with the speed decrease), it is determined that the mechanical pump is operating abnormally, and the process ends.

[0086] The key to the above mechanical pump status identification method lies in the extraction of the shutdown segment (only extract the speed decrease segment to ensure that there is no constant speed segment affecting the result). Specifically, for the extraction of the shutdown segment: when the main control status word C0 - Cn is triggered, the generator speed drops rapidly; select the segment with a speed range difference > n0 (such as 500 rpm), and after extracting the effective segment, use the Pearson correlation coefficient between the generator speed and the oil pressure at the inlet of the gearbox as the judgment of the mechanical pump status.

[0087] Specifically, when identifying the mechanical pump status, the Pearson correlation coefficient of 0.9 is used as the threshold; when the result is less than 0.9, it is regarded as the mechanical pump failure, otherwise it is effective.

[0088] This method comprehensively analyzes the operating status of the mechanical pump by gradually detecting the changes in the main control status word, the generator speed, and the oil pressure at the inlet of the gearbox during the shutdown process, as well as the changes in the high / low-speed electric pump status word, and finally determines whether the mechanical pump is operating normally or abnormally, or discards the data segments that do not meet the requirements.

[0089] S4. As Figure 4 shown, use the DBSCAN clustering algorithm to classify and identify the electric pump status. The specific steps are as follows:

[0090] S41. Start

[0091] The process starts, and the evaluation and judgment of the electric pump status begin.

[0092] S42. Exclude the oil pressure provided by the mechanical pump

[0093] Exclude the part of the oil pressure provided by the mechanical pump from the detection data to ensure that the subsequent data analysis is only for the working conditions of the electric pump.

[0094] S43. Use the Arrhenius formula to correct the oil pressure to 40°C

[0095] According to the Arrhenius formula, correct the oil pressure data to the state at the standard temperature (40°C) to eliminate the influence of temperature on the oil pressure data.

[0096] S44. Extract the data segments twice as long before and after the start time of the low / high-speed gear of the electric pump

[0097] Extract the data segments twice as long before and after the start time of the low or high-speed gear of the electric pump from the data for analyzing the operating characteristics of the electric pump at each stage before and after startup.

[0098] S45. Judgment condition 1: Whether the opening duration is greater than t and whether the average generator speed is greater than n0

[0099] Check whether the running opening duration of the electric pump is greater than the set threshold t, and at the same time detect whether the average speed of the generator is greater than the set value n0; if so (the condition is satisfied), go to the next step; if not (the condition is not satisfied), directly end the process, and the data segment does not meet the requirements and is discarded.

[0100] S46. Judgment condition 2: Whether the data segment duration is less than t1

[0101] Check whether the total duration of the data segment is less than the set threshold t1; if so (the duration is less than t1), directly end the process, and the data segment does not meet the requirements and is discarded; if not (the duration meets the requirements), go to the next step.

[0102] S47. Calculate the average increased oil pressure when the low / high gear is turned on

[0103] Calculate the average oil pressure increment during the opening period of the electric pump in the low or high gear, which is an important indicator for analyzing the operating state of the electric pump.

[0104] S48. Classify all units using the DBSCAN algorithm

[0105] Use the DBSCAN clustering algorithm to classify all data of the electric pump units and identify different types of operating states.

[0106] S49. Judgment condition 3: Whether this unit is an outlier

[0107] According to the DBSCAN classification result, judge whether the current electric pump unit belongs to an outlier. If so (classified as an outlier), determine that the electric pump is abnormal and the process ends; if not (classified normally), determine that the electric pump is normal and the process ends.

[0108] The key to the above electric pump state recognition method lies in the recognition of abnormal units. First, extract the average oil pressure provided by the electric pump in different gears, and then use the DBSCAN clustering algorithm to identify it. For abnormal units, the oil pressure provided by their electric pump in the low / high gear is usually abnormally lower / higher than that of normal units, which appears as noise in the DBSCAN algorithm.

[0109] Specifically, when recognizing the state of the electric pump, 10% of the total number of units in the whole field is used as the minimum number of points, and 20% of the effective value of the oil pressure provided by the high and low speed gears of the electric pump is used as the neighborhood radius for clustering.

[0110] This method accurately determines whether the operating state of the electric pump is normal or abnormal by eliminating the influence of the mechanical pump, correcting the oil pressure data, extracting key data segments, calculating the oil pressure increment, and performing clustering analysis on the data using the DBSCAN algorithm. This method effectively improves the accuracy of anomaly detection and ensures the reliability of the evaluation results.

[0111] The present invention analyzes the operating mechanism of the lubrication system of a wind turbine gearbox, the operating principles of the mechanical pump and the electric pump, derives the model of the lubrication system, and separates the state characteristics of the mechanical pump and the electric pump therefrom. Based on the existing sensor parameters in the lubrication system: main control state, generator speed, oil pressure at the inlet of the gearbox, oil temperature at the inlet of the gearbox and other measuring points, after extracting the effective segments according to the state recognition methods of the mechanical pump and the electric pump, the Pearson correlation coefficient and the DBSCAN algorithm are respectively used to process the segments, and the states of the mechanical pump and the electric pump are judged according to the algorithm results.

[0112] The method described in the present invention mainly involves the field of operating state monitoring and fault diagnosis of wind turbine units. When there is no dedicated sensor for monitoring the equipment, the data of other sensors in the system are used for state recognition, such as the state recognition method for the mechanical pump and the electric pump of the gearbox without sensors.

[0113] The present invention does not require additional sensors, which simplifies the structure of the monitoring system; traditional monitoring systems rely on a large number of sensors to collect equipment operation data, which not only increases the complexity and cost of the system, but is also limited by the lifespan and accuracy of the sensors. To save costs and simplify the structure of the monitoring system, and improve the reliability and practicality of the system, the present invention utilizes other monitoring variables in the gearbox lubrication system to achieve the state recognition of the mechanical pump and the electric pump without additional sensors.

[0114] The present invention improves the accuracy and reliability of the state recognition of the mechanical pump and the electric pump. Since the mechanism of the present invention is clear and the principles and processes of the analysis means are clear, the state recognition work of the mechanical pump and the electric pump can be carried out on a more scientific and rigorous basis. Through the detailed analysis of the operating mechanisms of the mechanical pump and the electric pump, the functional relationships between the oil pressure and the rotational speed of the mechanical pump, and between the oil temperature and the viscosity of the lubricating oil are established, and the oil pressure is decomposed accordingly. The states of the mechanical pump and the electric pump are successfully judged using the Pearson correlation coefficient and the DBSCAN clustering algorithm, and the abnormal units are accurately identified.

[0115] such as Figure 5As shown, DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is a density-based clustering algorithm mainly used to identify clustering structures and noise points in data. The algorithm determines whether a point is a core point by defining two parameters: ε (neighborhood radius) and MinPts (minimum number of points). Specifically, if a point contains at least MinPts points in its neighborhood, then the point is considered a core point, and the points closely related to it are grouped into the same cluster; in contrast, points with fewer than MinPts points in the neighborhood are considered border points or noise. The advantage of DBSCAN is that it can discover clusters of any shape and has good robustness to noisy data, without the need to pre-specify the number of clusters, which is very suitable for the clustering scenario of the pressure of an electric pump.

[0116] The present invention also discloses a computer program product, including a computer program that, when run by a processor, executes the steps of the method described above.

[0117] The present invention further discloses a computer-readable storage medium, on which a computer program is stored, and the computer program, when run by a processor, executes the steps of the method described above.

[0118] The present invention also discloses a state recognition system for a mechanical pump and an electric pump of a wind turbine gearbox, including a memory and a processor connected to each other, and a computer program is stored on the memory, and the computer program, when run by a processor, executes the steps of the method described above.

[0119] The products, media, and systems of the present invention, corresponding to the above method, also have the advantages described in the above method.

[0120] The implementation of all or part of the processes in the above-described embodiment methods of the present invention can also be completed by hardware related to computer program instructions. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable storage medium includes: any entity or device capable of carrying computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal, and software distribution medium, etc. The memory is used to store computer programs and / or modules. The processor realizes various functions by running or executing the computer programs and / or modules stored in the memory, and by calling the data stored in the memory. The memory can include high-speed random access memory, and can also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one magnetic disk storage device, flash device, or other volatile solid-state storage devices, etc.

[0121] The above is only the preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art in this technical field, several improvements and refinements made without departing from the principle of the present invention should be regarded as within the protection scope of the present invention.

Claims

1. A method for identifying the status of a mechanical pump and an electric pump in a wind turbine gearbox, characterized in that: Includes steps: S1. Construct a lubrication system operation model by extracting the gearbox inlet oil pressure, lubricating oil temperature and generator speed data; S2, decoupling the operating states of the gearbox mechanical pump and the electric pump based on the lubrication system operation model; S3, extract the shutdown fragment and use the Pearson correlation coefficient to identify the state of the mechanical pump; S4. Extract the data segments with twice the length before and after the electric pump is turned on at low speed or high speed, and use the DBSCAN clustering algorithm to classify and identify the electric pump status.

2. The method for identifying the status of a mechanical pump and an electric pump of a wind turbine gearbox according to claim 1, characterized in that: In step S1, the constructed lubrication system operation model is specifically: P=P E ·[1+ΔA(T)]+P M Where P is the gearbox inlet oil pressure, P E is the pressure provided by the electric pump at a constant temperature T0, ΔA(T) is the change in the reaction rate of the oil at temperature T, calculated using the Arrhenius equation, P M The pressure provided by the mechanical pump.

3. The method for identifying the status of a mechanical pump and an electric pump in a wind turbine gearbox according to claim 2, characterized in that: Decoupling of the mechanical pump operation state: When the electric pump state does not change and the gearbox cooling fan is not turned on, the oil pressure provided by the electric pump is relatively stable and the lubricating oil temperature does not change significantly. At this time, the change in the inlet oil pressure is only affected by the mechanical pump. The functional relationship between the inlet oil pressure and the mechanical pump can be solved, specifically: P M =k M n where k M is the conversion coefficient, and n is the generator speed.

4. The method for identifying the status of a mechanical pump and an electric pump of a wind turbine gearbox according to claim 3, characterized in that: Decoupling of the operating state of the electric pump: Using the wind wheel speed, combined with the functional relationship between the mechanical pump and the inlet oil pressure, the influence of the mechanical pump on the inlet oil pressure is eliminated; according to the Arrhenius equation, the influence of the lubricating oil temperature on the viscosity is corrected, and then the influence of the lubricating oil temperature on the inlet oil pressure is corrected; at this time, the change in oil pressure is only affected by the change in the state of the electric pump, and the oil pressure value that can be provided by each gear of the electric pump is solved, specifically: Among them, ΔA(T)=A(T)-A(T0); According to the Arrhenius equation: Among them, k W is the reaction rate constant, A is the frequency factor, E a is the activation energy, R is the gas constant, and T is the absolute temperature.

5. The method for identifying the status of a mechanical pump and an electric pump in a wind turbine gearbox according to any one of claims 1 to 4, characterized in that: The specific process of step S3 is: S31, start; S32, extracting the shutdown fragment; S33, determine whether the duration exceeds the set value t and whether the generator speed range is greater than n0; if so, proceed to the next step; S34, determine whether the high / low speed electric pump status word has changed; if so, proceed to the next step; S35, detecting the Pearson correlation coefficient between the generator speed and the gearbox inlet oil pressure; if the Pearson correlation coefficient is less than the set threshold, it is determined that the mechanical pump has failed, otherwise it is valid.

6. The method for identifying the status of a mechanical pump and an electric pump in a wind turbine gearbox according to claim 5, characterized in that: In step S32, the shutdown segment is: when the main control status word C0-Cn is triggered, the generator speed drops rapidly and the speed difference is >n0.

7. The method for identifying the status of a mechanical pump and an electric pump in a wind turbine gearbox according to any one of claims 1 to 4, characterized in that: The specific process of step S4 is: S41, Start S42, eliminating the oil pressure part provided by the mechanical pump from the detection data; S43, use the Arrhenius formula to correct the oil pressure to 40°C; S44, extracting twice the data fragments before and after the low / high gear opening period of the electric pump; S45, checking whether the operation start time of the electric pump is greater than the set threshold t, and at the same time detecting whether the average speed of the generator is greater than the set value n0; If yes, go to the next step; S46, check whether the total duration of the data segments is less than the set threshold t1; if not, proceed to the next step; S47, calculating the average oil pressure increment of the electric pump during the low speed or high speed gear opening period; S48. Based on the average oil pressure increment, the DBSCAN clustering algorithm is used to classify all electric pump unit data and identify different types of operating states; S49. According to the DBSCAN classification result, determine whether the current electric pump unit is an abnormal point; if so, determine that the electric pump is abnormal; if not, determine that the electric pump is normal.

8. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are performed.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the computer program performs the steps of the method according to any one of claims 1 to 7.

10. A computer device comprising a memory and a processor connected to each other, wherein a computer program is stored in the memory, characterized in that: When the computer program is executed by a processor, the computer program performs the steps of the method according to any one of claims 1 to 7.

Images