Fault diagnosis method and fault diagnosis system for wind driven generator

Abstract

The invention relates to a fault diagnosis method and a fault diagnosis system for a wind driven generator. The fault diagnosis method comprises the steps that a physical model is established, and the physical model reflects complete physical structure characteristics of a shafting; collecting a series of strain signals of the shafting at a series of time points; calculating a series of load distributions of the wind generator at the series of time points through a physical model based on the series of strain signals; calculating a historical load distribution of the wind driven generator based on the series of load distributions; and obtaining operation state parameters of the wind driven generator based on historical load distribution analysis. The fault diagnosis method and the fault diagnosis system are convenient for monitoring the operation state of the wind driven generator.

Description

Technical Field

[0001] The present invention relates to the technical field of wind power generation, and in particular to a fault diagnosis method and a fault diagnosis system for a wind power generator. Background Art

[0002] In recent years, with the continuous development and application of large wind turbines, especially offshore wind turbines, their reliability and practicality have attracted more and more attention. Component failure of wind turbines will cause power production loss and increase maintenance costs, thus affecting operating benefits. Therefore, the operation status monitoring technology of wind turbines has received more and more attention.

[0003] Patent document 201580049451.7 discloses a monitoring method for obtaining bearing load based on eddy current displacement sensor, which establishes a mathematical model based on statics to obtain the relationship between the displacement value and the actual load state. Patent document 202211164008.2 discloses a wind turbine monitoring system and method based on simulation algorithm, which combines SCADA data with a training model based on dynamic modeling of wind turbines to predict the tower state. Patent document WO2011 / 081085 discloses an abnormality diagnosis device for rolling bearings and wind turbines, which uses a vibration sensor to measure the RMS value of the vibration waveform, and uses the AC component of the vibration waveform to detect the abnormal state of the bearing.

[0004] Most of these existing monitoring systems detect faults based on vibration or acoustic principles. These signals are easily interfered by environmental noise and require special algorithms to extract signal features. At the same time, these signals are not directly related to the nature of the faults caused by material fatigue. Moreover, this method focuses on the status indication after the fault occurs, and is not suitable for monitoring the early stages of the fault and taking early decisions to extend the service life. In addition, since the extreme conditions of wind turbines vary with wind conditions, size and type of components, the conditions that trigger alarms are not standard. Summary of the invention

[0005] Therefore, the technical problem to be solved by the present invention is to provide an improved fault diagnosis method and fault diagnosis system for a wind turbine.

[0006] The above technical problems are solved by a fault diagnosis method for a wind turbine according to the present invention. The wind turbine includes a shaft system consisting of a main shaft, a main bearing supporting the main shaft, and a bearing seat supporting the main bearing. The fault diagnosis method includes: establishing a physical model, wherein the physical model reflects the complete physical structural characteristics of the shaft system; collecting a series of strain signals of the shaft system at a series of time points; calculating a series of load distributions of the wind turbine at the series of time points through the physical model based on the series of strain signals; calculating the historical load distribution of the wind turbine based on the series of load distributions; and obtaining the operating status parameters of the wind turbine based on the analysis of the historical load distribution. This method calculates the load distribution of the wind turbine based on the strain signal, and analyzes and predicts the operating status of the wind turbine based on the historical load distribution. The analysis result is less affected by environmental conditions and has high accuracy.

[0007] According to a preferred embodiment of the present invention, the series of strain signals can be collected from the bearing seat. As a key force-bearing component of the shaft system, the bearing seat can provide reliable strain signals and is easy to set up sensors.

[0008] According to another preferred embodiment of the present invention, the physical model can be established by finite element analysis or by advanced bearing analysis tools. These modeling methods are simple and effective.

[0009] According to another preferred embodiment of the present invention, the complete physical structural features of the shaft system reflected by the physical model may include the complete geometric features and complete material features of the shaft system. The load distribution state of the shaft system can be accurately calculated by reflecting the complete physical structural features of the shaft system.

[0010] According to another preferred embodiment of the present invention, the operating state parameter may include one or more of the following parameters of one or more target components of the wind turbine: expected remaining life, structural strength and / or wear state. Preferably, the one or more target components of the wind turbine may include at least a main bearing. In addition, the target component may also include a bearing seat and a main shaft, etc.

[0011] According to another preferred embodiment of the present invention, calculating the historical load distribution of the wind turbine based on the series of load distributions may include: calculating the historical load distribution of the wind turbine based on the series of load distributions and considering relevant data in a trained database. The relevant data in the trained database may help calculate the historical load distribution more accurately.

[0012] According to another preferred embodiment of the present invention, the fault diagnosis method may further include: when the analyzed operating state parameters of the wind turbine generator correspond to predetermined risk consequences, issuing a warning signal to instruct relevant staff to prevent and handle relevant faults as soon as possible.

[0013] The above technical problem is also solved by a fault diagnosis system for a wind turbine according to the present invention. The fault diagnosis system is used in the above fault diagnosis method, and the fault diagnosis system includes: a model building module, which is configured to build a physical model of the complete physical structural characteristics of the reaction shaft system; a signal acquisition module, which includes one or more strain sensors configured to collect a series of strain signals of the shaft system at a series of time points; a real-time load distribution calculation module, which is configured to calculate a series of load distributions of the wind turbine at a series of time points through a physical model based on the series of strain signals; a historical load distribution calculation module, which is configured to calculate the historical load distribution of the wind turbine based on the series of load distributions; and an analysis module, which is configured to obtain the operating status parameters of the wind turbine based on the historical load distribution analysis. This system calculates the load distribution of the wind turbine based on the strain signal, and analyzes and predicts the operating status of the wind turbine based on the historical load distribution. The analysis result is less affected by environmental conditions and has high accuracy.

[0014] According to a preferred embodiment of the present invention, the one or more strain sensors may be arranged on the bearing seat. As a key force-bearing component of the shaft system, the bearing seat can provide reliable strain signals and is easy to set sensors. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present invention is further described below in conjunction with the accompanying drawings. The same reference numerals in the drawings represent elements with the same functions.

[0016] Figure 1 A flowchart showing a fault diagnosis method according to an exemplary embodiment of the present invention; and

[0017] Figure 2 A schematic diagram showing installation positions of sensors of a fault diagnosis system according to an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0018] The following will describe the specific implementation of the fault diagnosis method and fault diagnosis system for wind turbines according to the present invention in conjunction with the accompanying drawings. The following detailed description and the accompanying drawings are used to exemplarily illustrate the principles of the present invention. The present invention is not limited to the preferred embodiments described, and the protection scope of the present invention is defined by the claims.

[0019] According to an embodiment of the present invention, a fault diagnosis method for a wind turbine is provided, which analyzes the operating state of the wind turbine based on a complete physical model and a true strain signal of the wind turbine.

[0020] According to practical experience, the main cause of wind turbine failure is usually material fatigue, which is usually caused by the accumulation of loads applied to the structural components of the wind turbine. The main components that affect the load distribution of the wind turbine are the components in the shaft system. The shaft system of the wind turbine consists of a main shaft 1, a main bearing 2 and a bearing seat 3, wherein the main shaft 1 is supported on the main bearing 2, and the main bearing 2 is supported on the bearing seat 3. The main shaft 1 transfers the load from the blades to the main bearing 2, and then to the bearing seat 3. The bearing seat 3 is connected to the tower of the wind turbine. When the load is applied to the shaft system, the load will be transferred to various components of the wind turbine, so that all mechanical components of the wind turbine will be deformed. The elastic or deformation state of the main shaft 1 and the bearing seat 3 will affect the distribution of the load. Based on the above principle, the fault diagnosis method according to the present invention can obtain the load distribution of the wind turbine by detecting the strain signal of the shaft system, and then analyze the operating state of the wind turbine.

[0021] Combine the following Figure 1 To illustrate the specific content of the fault diagnosis method of the present invention. Figure 1 FIG. 2 shows a flow chart of a fault diagnosis method according to an exemplary embodiment of the present invention. Figure 1 As shown, the fault diagnosis method mainly includes the following steps.

[0022] First, in step S1, a physical model is established. The physical model here is a physical model that reflects the complete physical structural characteristics of the shaft system of the wind turbine. This physical model is established in the calculation module of the computing device. On the one hand, the complete physical structural characteristics of the shaft system include the complete geometric characteristics of the shaft system, including all shape characteristics and all dimensional characteristics of all components of the shaft system. On the other hand, the complete physical structural characteristics of the shaft system also include the complete material characteristics of the shaft system, including the material stiffness characteristics of all components of the shaft system, etc. This complete physical model can reflect the correlation between strain and load distribution state, and can therefore be used to accurately calculate the load distribution on the entire shaft system through strain.

[0023] The physical model in step S1 can be established according to various known methods. For example, the physical model can be established by finite element analysis (FEA). Finite element analysis is a basic mechanical analysis modeling method in the art, and its specific process is well known and will not be repeated here. Alternatively, it can also be established by some advanced bearing analysis tools currently available on the market. These advanced bearing analysis tools are usually analysis software developed by bearing manufacturers, such as the Bearinx tool of the applicant. An effective physical model of the shaft system can be established by these modeling methods or tools.

[0024] Next, in step S2, data collection is performed. Specifically, in step S2, a series of strain signals of the shaft system at a series of time points are collected. The required strain signals are collected by one or more strain sensors 4 arranged on the corresponding components. Figure 2 A schematic diagram of the shaft system is shown, such as Figure 2 As shown, the strain sensor 4 can preferably be arranged on the bearing seat 3 to directly collect the strain signal from the bearing seat 3. As a relatively stationary component in the shaft system and arranged on the outside, the bearing seat 3 is convenient for installing the strain sensor 4. However, as needed, the strain sensor 3 can also be arranged on other components in the shaft system. The number of strain sensors 4 can be set as needed.

[0025] The strain sensor 4 collects a series of strain signals corresponding to the shaft system at a series of predetermined time points (at least two). These time points can be allocated according to a predetermined time plan, for example, they can be allocated according to a predetermined time interval. The specific number of time points and the allocation interval can be designed based on factors such as practical experience and specific needs.

[0026] The strain sensor 4 completes the collection of a strain signal at each predetermined time point, thereby obtaining a corresponding strain signal. Each of a series of time points corresponds to a unique strain signal, thereby obtaining a series of strain signals. These strain signals can be transmitted to a related data processing device for further processing and analysis. The transmission of the strain signal can be carried out by various well-known wired or wireless methods. The data processing device can be a computer located locally or remotely.

[0027] After the strain signal is obtained, step S3 is performed. In step S3, based on the obtained series of strain signals, a series of load distributions of the wind turbine at a corresponding series of time points are calculated by the established physical model. The load distribution at the corresponding time point can be calculated based on each strain signal, thereby obtaining a plurality of load distributions corresponding to a series of time points.

[0028] Next, in step S4, the historical load distribution of the wind turbine is calculated based on the series of load distributions obtained. The historical load distribution reflects the process of the load distribution of the wind turbine changing over time. Preferably, the process of calculating the historical load distribution of the wind turbine not only considers the load distribution based on the sensor measurement signal, but also considers the relevant data in the trained database. This is a process of machine learning based on existing data information. Through this machine learning process, the obtained historical load distribution can be made more accurate.

[0029] After obtaining the historical load distribution, in step S5, the operating state parameters of the wind turbine are obtained based on the obtained historical load distribution analysis. This method of analyzing the operating state through historical load distribution is known in the art and can be performed based on practical experience and established databases, which will not be described in detail here.

[0030] By analyzing the historical load distribution, the operating status of each component of the wind turbine can be understood. One or more key components of interest can be selected as analysis objects, and their corresponding operating status parameters can be obtained. The selected analysis object can be called a target component. The operating status parameters of the target component can include one or more of the expected remaining life, structural strength and / or wear status. The target component can preferably include at least the main bearing 2, and can also include other components, such as the main shaft 1 in the shaft system, the bearing seat 3 or components outside the shaft system.

[0031] Preferably, the fault diagnosis method may further include an additional step S6. In step S6, when the analyzed operating state parameter corresponds to a predetermined risk consequence, an early warning signal is issued. For example, when the expected remaining life or structural strength of one or more target components is lower than a predetermined value, or when the degree of wear of one or more target components is greater than a predetermined value, it can be determined that the operating state parameter corresponds to a predetermined risk consequence. At this time, a corresponding early warning signal can be issued to the relevant operator or manager so as to deal with the relevant risk in a timely manner.

[0032] Corresponding to the above fault diagnosis method, an embodiment of the present invention further provides a fault diagnosis system for a wind turbine generator, which is used to perform the above fault diagnosis method. The fault diagnosis system includes: a model building module, a signal acquisition module, a real-time load distribution calculation module, a historical load distribution calculation module and an analysis module.

[0033] The model building module is used to execute step S1, and is configured to build a physical model of the complete physical structural characteristics of the reaction shaft system.

[0034] The signal acquisition module is used to perform step S2, and includes one or more strain sensors 4 configured to acquire a series of strain signals corresponding to the shaft system at a series of time points. As mentioned above, these strain sensors 4 can preferably be arranged on the bearing seat 3.

[0035] The real-time load distribution calculation module is used to execute step S3, and is configured to calculate a series of load distributions of the wind turbine at a corresponding series of time points through the above physical model based on the collected series of strain signals.

[0036] The historical load distribution calculation module is used to perform step S4, and is configured to calculate the historical load distribution of the wind turbine based on the obtained series of load distributions. As mentioned above, the historical load distribution calculation module preferably also considers the relevant data in the trained database to calculate the historical load distribution of the wind turbine.

[0037] The analysis module is used to perform step S5, and is configured to obtain the operating state parameters of the wind turbine based on the historical load distribution analysis. The operating state parameters may include one or more of the expected remaining life, structural strength and / or wear state of the target component.

[0038] In addition, the fault diagnosis system can also be used to execute the early warning module of the additional step S6, which is configured to issue an early warning signal when the analyzed operating status parameter corresponds to a predetermined risk consequence.

[0039] In the fault diagnosis system, the model building module, the real-time load distribution calculation module, the historical load distribution calculation module, the analysis module and the early warning module can be virtual modules integrated on a local or remote data processing device, and the signal acquisition module is a physical module composed of sensors, and the virtual module and the physical module can be communicatively connected.

[0040] The fault diagnosis method and fault diagnosis system according to the present invention calculate and analyze the load distribution of the wind turbine based on the physical model and strain signal of the shaft system of the wind turbine, thereby accurately monitoring the operating state of the wind turbine and predicting related faults. This is conducive to early processing and prevention of wind turbine faults, thereby reducing losses. At the same time, the fault diagnosis method and fault diagnosis system are easy to implement and low cost.

[0041] Although possible embodiments are described by way of example in the above description, it should be understood that there are still a large number of variations of embodiments through the combination of all known and other technical features and embodiments that are easily conceivable to the skilled person. It should also be understood that the exemplary embodiment is only an example and that such an embodiment in no way limits the scope of protection, application and configuration of the present invention. The above description is more to provide the skilled person with a technical guide for converting at least one exemplary embodiment, wherein various changes can be made, especially changes in the functions and structures of the components, as long as they do not depart from the scope of protection of the claims.

[0042] Reference numerals list

[0043] 1 Spindle

[0044] 2 Main bearing

[0045] 3 Bearing seat

[0046] 4 Strain Sensors

Claims

1. A fault diagnosis method for a wind turbine, wherein the wind turbine comprises a shaft system consisting of a main shaft (1), a main bearing (2) supporting the main shaft (1), and a bearing seat (3) supporting the main bearing (2), wherein: The fault diagnosis method comprises: Establishing a physical model, wherein the physical model reflects the complete physical structural characteristics of the shaft system; Collecting a series of strain signals of the shaft system at a series of time points; Calculating a series of load distributions of the wind turbine at a series of time points through the physical model based on the series of strain signals; Calculating a historical load distribution of the wind turbine based on the series of load distributions; The operating status parameters of the wind turbine are obtained based on the historical load distribution analysis.

2. The fault diagnosis method according to claim 1, characterized in that: The series of strain signals are collected from the bearing seat (3).

3. The fault diagnosis method according to claim 1, characterized in that: The physical model is established by finite element analysis or by advanced bearing analysis tools.

4. The fault diagnosis method according to claim 1, characterized in that: The complete physical structural characteristics of the shaft system reflected by the physical model include the complete geometric characteristics and complete material characteristics of the shaft system.

5. The fault diagnosis method according to claim 1, characterized in that: The operating status parameter includes one or more of the following parameters of one or more target components of the wind turbine: expected remaining life, structural strength and / or wear status.

6. The fault diagnosis method according to claim 5, characterized in that: The one or more target components of the wind turbine include at least the main bearing (2).

7. The fault diagnosis method according to claim 1, characterized in that: Calculating the historical load distribution of the wind turbine based on the series of load distributions includes calculating the historical load distribution of the wind turbine based on the series of load distributions and considering relevant data in a trained database.

8. The fault diagnosis method according to any one of claims 1 to 7, characterized in that: The fault diagnosis method further comprises: when the operating state parameter of the wind turbine obtained by analysis corresponds to a predetermined risk consequence, issuing a warning signal.

9. A fault diagnosis system for a wind turbine, used to execute the fault diagnosis method according to any one of claims 1 to 8, characterized in that: The fault diagnosis system comprises: A model building module, configured to build a physical model reflecting the complete physical structural characteristics of the shaft system; A signal acquisition module, comprising one or more strain sensors (4) configured to acquire a series of strain signals of the shaft system at a series of time points; A real-time load distribution calculation module, configured to calculate a series of load distributions of the wind turbine at a series of time points based on the series of strain signals through the physical model; a historical load distribution calculation module configured to calculate a historical load distribution of the wind turbine based on the series of load distributions; and An analysis module is configured to obtain operating status parameters of the wind turbine based on the historical load distribution analysis.

10. The fault diagnosis system according to claim 9, characterized in that: The one or more strain sensors (4) are arranged on the bearing seat (3).

Images