Fan variable pitch bearing multi-direction vibration collaborative fault identification method and system

Abstract

The present application belongs to the technical field of wind power equipment state monitoring and fault diagnosis, and particularly relates to a wind turbine variable-pitch bearing multi-direction vibration cooperative fault identification method and system. The method comprises: collecting vibration signals obtained along different directions at the same structure measuring point during the operation of the wind turbine variable-pitch bearing, to obtain multi-direction vibration data with time consistency; based on the multi-direction vibration data, calculating the synchronous correlation between vibration responses in different directions in the frequency domain to form fusion vibration information reflecting the common excitation effect; generating multi-frequency joint statistical results according to the joint changes between different frequency components in the fusion vibration information, and judging the operation state of the wind turbine variable-pitch bearing according to the change characteristics of the multi-frequency joint statistical results to realize fault identification of the wind turbine variable-pitch bearing. The present application has strong noise resistance and engineering applicability.

Description

Technical Field

[0001] This invention belongs to the field of wind power equipment condition monitoring and fault diagnosis technology, specifically relating to a method and system for identifying multi-directional vibration collaborative faults in wind turbine pitch bearings. Background Technology

[0002] As a critical load-bearing and pitch-drive component in the wind turbine operating system, the operating condition of the wind turbine directly affects the safety and power generation efficiency of the wind turbine. During long-term service, the pitch bearing structure must continuously withstand multiple effects, including aerodynamic loads, gravitational loads, pitch drive torque, and wind speed fluctuations. Its internal raceways, rolling elements, cages, and other critical components are prone to fatigue damage, wear, or localized defects, leading to abnormal vibrations. If these abnormal vibrations are not identified in a timely manner, the fault may further escalate, ultimately affecting the safe operation of the wind turbine.

[0003] Due to the complexity of pitch bearing structures and the diversity of vibration propagation paths, the vibration response of pitch bearings typically exhibits significant direction-dependent characteristics. The same fault excitation can produce vibration responses with different amplitude distributions, frequency weights, and energy concentration regions in different structural directions. Furthermore, fault features are often distributed in a modulated form across multiple frequency components, accompanied by strong background noise interference, resulting in a low signal-to-noise ratio for the fault signal. This poses a significant challenge to stable and reliable fault feature extraction.

[0004] Existing technologies for fault identification of pitch bearing vibration signals mainly focus on time-domain, frequency-domain, and time-frequency-domain signal analysis methods. For example, by analyzing the periodic impacts, spectral characteristics, or energy distribution in the vibration signal, the operating status of key components of the pitch bearing can be determined. These methods are effective when fault characteristics are obvious and operating conditions are relatively stable, but they are prone to problems such as feature ambiguity and unstable identification under strong noise backgrounds or weak modulation conditions.

[0005] To enhance sensitivity to weak fault characteristics, some technical solutions introduce more complex frequency domain analysis strategies. By characterizing the correlation between different frequency components in the vibration signal, they aim to improve the response to fault modulation features. However, most of these methods are based on the analysis of vibration signals in a single direction, failing to fully consider the physical correlation between vibration responses in different directions at the same measurement point. In practical engineering, the vibration of a pitch bearing structure is essentially a projection of a spatial vibration vector in different directions. A signal in a single direction can only reflect the local response of the fault excitation, easily leading to incomplete fault information and affecting the stability and reliability of the identification results.

[0006] To address the aforementioned issues, some studies have attempted to introduce multi-channel or multi-sensor vibration signals for joint analysis to enhance the ability to represent fault characteristics by utilizing the correlation between different signals. Existing multi-channel methods primarily focus on information fusion at the feature layer or decision layer, essentially mining the statistical correlation between signals from different channels through a data-driven approach. However, they lack direct modeling of the multi-frequency coordinated variation relationships and their inherent constraints prevalent in modulation faults. From a physical mechanism perspective, faults in the rotating components of wind turbine pitch bearings typically manifest as coordinated changes between impact excitation and carrier frequency vibration. Their characteristic information is distributed across multiple frequency components and their interrelationships. Relying solely on a single frequency or simple correlation analysis is insufficient to comprehensively characterize these complex characteristics.

[0007] Furthermore, some model- or learning-based methods often rely on a large amount of sample data under different operating conditions and fault states during implementation, which can lead to problems such as implementation complexity, insufficient interpretability, or limited engineering applicability. Under complex operating conditions and weak fault conditions, the recognition results may still be unstable or the generalization ability may decrease.

[0008] Therefore, there is an urgent need for a fault identification method for pitch bearings of wind turbines. Based on the synchronous acquisition of vibration signals in multiple directions, this method should make full use of the synchronous correlation between vibration responses in different directions and further introduce a statistical constraint mechanism that can reflect the characteristics of multi-frequency coordinated changes. This will enable the stable extraction and reliable identification of weak fault features of pitch bearings, thereby improving the accuracy and engineering applicability of pitch bearing operation status monitoring. Summary of the Invention

[0009] To address the problems of existing wind turbine pitch bearing fault identification methods that rely heavily on single-directional vibration signals, fail to fully utilize the physical correlation between multi-directional vibration data, and exhibit poor characteristic stability under weak fault and strong noise conditions, this invention provides a multi-directional vibration collaborative fault identification method and system for wind turbine pitch bearings. The aim is to achieve effective enhancement and reliable identification of pitch bearing fault characteristics through synchronous acquisition and collaborative processing of multi-directional vibration signals, combined with multi-frequency joint statistical analysis.

[0010] To achieve the above objectives, the present invention provides the following solution: A method for collaborative fault identification of multi-directional vibration in wind turbine pitch bearings, the method comprising: Time-aligned acquisition of vibration signals obtained from the same structural measuring point along different directions during the operation of the wind turbine pitch bearing was performed to obtain multi-directional vibration data with time consistency. Based on the multi-directional vibration data, the synchronous correlation between vibration responses in different directions in the frequency domain is calculated to form fused vibration information reflecting the common excitation effect; Based on the joint changes between different frequency components in the fused vibration information, multi-frequency joint statistical results are generated, and the operating status of the wind turbine pitch bearing is judged according to the change characteristics of the multi-frequency joint statistical results, thereby realizing fault identification of the wind turbine pitch bearing.

[0011] Preferably, a method for time-aligned acquisition of vibration signals obtained at the same structural measuring point along different directions during the operation of a wind turbine pitch bearing to obtain multi-directional vibration data with time consistency includes: Multiple directionally sensitive vibration acquisition units are installed at the same physical location on the pitch bearing of the wind turbine. Based on the vibration acquisition unit, the vibration signal of the wind turbine pitch bearing is acquired in real time and synchronously in different directions to obtain multi-directional vibration data with time consistency.

[0012] Preferably, the method for calculating the synchronous correlation between vibration responses in different directions in the frequency domain based on the multi-directional vibration data to form fused vibration information reflecting the common excitation effect includes: The multi-directional vibration data are subjected to Fourier transform to obtain a frequency domain representation; Based on the frequency domain representation, the synchronous correlation between vibration responses in different directions is calculated in the frequency domain to obtain fused vibration information that reflects the common excitation effect in multi-directional vibration data.

[0013] Preferably, the method for calculating the synchronous correlation between vibration responses in different directions in the frequency domain based on the frequency domain representation to obtain fused vibration information reflecting the common excitation effect in multi-directional vibration data includes: ; in, This indicates that the multi-directional vibration data sequence is at a frequency The result of the synchronization consistency measurement at the location, i.e., the fused vibration information; and These represent the amplitude coefficients of the vibration response in different directions; superscript Indicates the complex conjugate symbol; This represents the equivalent frequency response of the structure to fault excitation. This represents the equivalent weighting factor introduced by the structural transitivity. Represents the mathematical expectation; and These represent the first and second digits of the multi-directional vibration data sequence, respectively. Next and first The arrival time of the secondary fault impact It represents the imaginary unit.

[0014] Preferably, the method for generating multi-frequency joint statistical results based on the joint changes between different frequency components in the fused vibration information includes: By introducing a frequency offset parameter, combined statistical analysis is performed on the fused vibration information at multiple frequency locations to generate multi-frequency joint statistical results.

[0015] Preferably, the method for introducing a frequency offset parameter and combining and statistically analyzing the fused vibration information at multiple frequency locations to generate multi-frequency joint statistical results includes: ; in, This indicates that the fused vibration information is present at the frequency. and frequency offset parameters Multi-frequency joint statistical results under certain conditions.

[0016] Preferably, the method for determining the operating status of the wind turbine pitch bearing based on the variation characteristics of the multi-frequency joint statistical results, thereby achieving fault identification of the wind turbine pitch bearing, includes: Based on the variation amplitude of the multi-frequency joint statistical results at a predetermined frequency combination, a judgment is made on whether there is abnormal vibration behavior in the wind turbine pitch bearing, thereby realizing the fault identification of the wind turbine pitch bearing.

[0017] The present invention also provides a multi-directional vibration collaborative fault identification system for wind turbine pitch bearings. The system is used to implement the aforementioned multi-directional vibration collaborative fault identification method for wind turbine pitch bearings. The system includes: a data acquisition module, a correlation analysis module, and a fault identification module. The data acquisition module is used to time-align the vibration signals obtained at the same structural measuring point along different directions during the operation of the wind turbine pitch bearing, so as to obtain multi-directional vibration data with time consistency. The correlation analysis module is used to calculate the synchronous correlation between vibration responses in different directions in the frequency domain based on the multi-directional vibration data, forming fused vibration information that reflects the common excitation effect; The fault identification module is used to generate multi-frequency joint statistical results based on the joint changes between different frequency components in the fused vibration information, and to judge the operating status of the wind turbine pitch bearing based on the change characteristics of the multi-frequency joint statistical results, thereby realizing fault identification of the wind turbine pitch bearing.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention proposes a method and system for collaborative fault identification of multi-directional vibration in wind turbine pitch bearings. The core of this method lies in the collaborative utilization mechanism of multi-directional vibration data. This mechanism simultaneously acquires vibration signals from different directions at the same measuring point, fully utilizing the physical characteristics of multi-directional vibration data in terms of temporal consistency and frequency complementarity, thus providing a reliable data foundation for subsequent information fusion and fault analysis.

[0019] Based on this, the present invention analyzes the synchronous correlation between multi-directional vibration signals, constructs fused vibration information that reflects the common excitation effect, effectively highlights the synchronous vibration component caused by the same fault excitation, and weakens the influence of uncorrelated vibration and random noise on the analysis results, thereby improving the stability and reliability of the analysis.

[0020] Furthermore, this invention introduces multi-frequency joint statistical analysis to characterize the coordinated changes among different frequency components in the fused vibration information, directly reflecting the consistency changes in multi-frequency responses caused by fault excitation. Compared to analysis methods that rely solely on a single frequency or low-order statistical features, this approach can effectively enhance fault-related features and improve the ability to identify abnormal vibration behavior in pitch bearings even under weak modulation and strong noise conditions.

[0021] Furthermore, this invention starts from the physical mechanism of vibration response, and performs collaborative processing and statistical analysis of multi-directional vibration information. The processing flow is clear, the characteristic meanings are well-defined, and it has good engineering interpretability. This method does not rely on complex model training or a large number of prior samples, and is applicable to the condition monitoring and fault identification of wind turbine pitch bearings under complex operating conditions, possessing strong engineering applicability and promotional value. Attached Figure Description

[0022] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart of a multi-directional vibration information collaborative fault identification method for wind turbine pitch bearings according to an embodiment of the present invention. Figure 2 This is a schematic diagram of a faulty bearing according to an embodiment of the present invention; Figure 3 The diagram shows the waveform, spectrum, and envelope spectrum of the horizontal vibration signal according to an embodiment of the present invention, wherein (a) is a waveform diagram, (b) is a spectrum diagram, and (c) is an envelope spectrum diagram. Figure 4The following are schematic diagrams of the waveform, spectrum, and envelope spectrum of the vertical vibration signal according to an embodiment of the present invention, wherein (a) is a waveform diagram, (b) is a spectrum diagram, and (c) is an envelope spectrum diagram; Figure 5 This is a schematic diagram of the multi-frequency joint statistical results of an embodiment of the present invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0026] Example 1 This invention provides a method for multi-directional vibration collaborative fault identification of wind turbine pitch bearings, including: Time-aligned acquisition of vibration signals obtained from the same structural measuring point along different directions during the operation of the wind turbine pitch bearing was performed to obtain multi-directional vibration data with time consistency. Based on the multi-directional vibration data, the synchronous correlation between vibration responses in different directions in the frequency domain is calculated to form fused vibration information reflecting the common excitation effect; Based on the joint changes between different frequency components in the fused vibration information, multi-frequency joint statistical results are generated, and the operating status of the wind turbine pitch bearing is judged according to the change characteristics of the multi-frequency joint statistical results, thereby realizing fault identification of the wind turbine pitch bearing.

[0027] The specific implementation process of this invention is as follows: This invention provides a method for multi-directional vibration collaborative fault identification of wind turbine pitch bearings. First, vibration signals from different directions at the same measuring point on the wind turbine pitch bearing are synchronously acquired to obtain multi-directional vibration data with temporal consistency. Then, utilizing the coherence and phase synchronization characteristics of the multi-directional vibration data in the frequency domain, a fused vibration information reflecting the common excitation effect is constructed to highlight the synchronous vibration component caused by fault excitation and suppress directional uncorrelated interference. Finally, multi-frequency joint statistical analysis is introduced to characterize the collaborative variation characteristics between different frequency components in the fused vibration information, thereby effectively enhancing the identifiability of weak fault modulation features. This method fully utilizes the inherent correlation characteristics of multi-directional vibration information, without relying on complex models or a large number of prior samples. It maintains stable fault identification performance even under complex operating conditions and strong noise backgrounds. It is suitable for engineering condition monitoring and fault identification of wind turbine pitch bearings, and for analyzing vibration anomalies in structural parts of wind turbine pitch bearings subjected to aerodynamic excitation.

[0028] Figure 1 This is a flowchart of a multi-directional vibration information collaborative fault identification method for wind turbine pitch bearings according to an embodiment of the present invention. Figure 1 As shown, the wind turbine pitch bearing fault identification method of this embodiment includes: S1. Time-aligned acquisition of vibration signals obtained from the same structural measuring point along different directions during the operation of the wind turbine pitch bearing, to obtain multi-directional vibration data with time consistency. Specifically, this includes: Multiple directionally sensitive vibration acquisition units are set up at the same physical location of the wind turbine pitch bearing, that is, vibration acceleration sensors and industrial-grade data acquisition cards are installed at the measuring points of the wind turbine pitch bearing housing to ensure that vibration responses in different directions have a consistent time reference under fault excitation. Based on the vibration acquisition unit, the vibration signal of the wind turbine pitch bearing is collected in real time and synchronously in different directions. This ensures that the vibration signals in different directions maintain a consistent time reference under fault excitation, thereby obtaining multi-directional vibration data with time consistency. The vibration signal is a vibration acceleration signal.

[0029] This step involves synchronously sensing the vibration response generated at the same structural measuring point along multiple spatial directions during the operation of the wind turbine pitch bearing. By using a unified time reference to perform time-series registration of the vibration responses in different directions, spatiotemporal consistent vibration data with the same excitation source, the same acquisition time, and multi-directional response characteristics is formed, i.e., multi-directional vibration data with time consistency.

[0030] Multi-directional vibration data includes vibration signals acquired along at least the horizontal and vertical directions.

[0031] S2. Utilizing the characteristics of multi-directional vibration data in terms of frequency domain correlation and phase synchronization, common frequency domain response components are extracted to construct fused vibration information reflecting the common excitation effect. Specifically, this includes: The multi-directional vibration data are subjected to Fourier transform to obtain a frequency domain representation; Based on the frequency domain representation of the multi-directional vibration data, the common frequency domain response components are extracted to obtain fused vibration information that reflects the common excitation effect.

[0032] This step is based on the multi-directional vibration data obtained in step S1. According to the coherence and phase synchronization characteristics (synchronization correlation) of the vibration response in each direction in the frequency domain, the common frequency domain response components are extracted to form fused vibration information that reflects the common excitation effect, so as to describe the coordinated change law of multi-directional response under the same excitation.

[0033] In practice, after synchronously acquiring vibration signals from multiple directions, Fourier transforms are first performed on the vibration signals from different directions to obtain their frequency domain representation. Taking two vibration signals acquired from the same measuring point along different directions as an example, their time domain expression can be expressed as: in, and These represent vibration signals in different directions; and Represents the unit direction vector in the corresponding direction; Indicates the first The intensity of the secondary failure impact; Characteristic coefficients related to structural state; The equivalent structural response function induced by the fault impact; Indicates the first The arrival time of the secondary fault impact; and This represents the noise term in the vibration signal in each direction. Superscript indicates a time variable. This indicates transpose.

[0034] Based on the time-domain expression, a Fourier transform of the multi-directional vibration data can be performed to obtain its frequency-domain representation: in, and These represent vibration signals in different directions at different frequencies. Frequency domain representation at; and Amplitude coefficients representing vibration responses in different directions; This represents the equivalent frequency response of the structure to fault excitation. and This represents the noise component in the frequency domain; It represents the imaginary unit.

[0035] Based on the above frequency domain representation, fused vibration information reflecting the common excitation effect is constructed, which is defined as: Substituting the frequency domain representation and expanding it, we have: in, This indicates that the multi-directional vibration data sequence is at a frequency The result of the synchronization consistency measurement at the location, i.e., the fused vibration information; and These represent the amplitude coefficients of the vibration response in different directions; superscript Indicates the complex conjugate symbol; This represents the equivalent frequency response of the structure to fault excitation. This represents the equivalent weighting factor introduced by the structural transitivity. Represents the mathematical expectation; and These represent the first and second digits of the multi-directional vibration data sequence, respectively. Next and first The arrival time of the secondary fault impact.

[0036] The following relationship is satisfied during the derivation process: Among them, superscript Indicates the complex conjugate symbol. Represents the mathematical expectation. and This represents the noise component in the frequency domain.

[0037] The above relationship indicates that, due to the lack of correlation between noise terms in different directions, the noise cross term can be ignored under the statistical average meaning, thus making the fused vibration information mainly reflect the synchronous vibration component caused by the same fault excitation.

[0038] By constructing fused vibration information This method can highlight the synchronous response caused by a common fault source in multi-directional vibration signals under second-order statistical significance, and effectively suppress direction-independent random noise and interference components. The processing result provides a fused input signal with a high signal-to-noise ratio for subsequent fault feature extraction based on multi-frequency joint statistics.

[0039] S3. Based on the joint changes between different frequency components in the fused vibration information, generate multi-frequency joint statistical results, and determine the operating status of the wind turbine pitch bearing based on the changing characteristics of the multi-frequency joint statistical results, thereby achieving fault identification of the wind turbine pitch bearing. Specifically, this includes: By introducing a frequency offset parameter, combined statistical analysis is performed on the fused vibration information at multiple frequency locations to generate multi-frequency joint statistical results. Based on the variation amplitude of the multi-frequency joint statistical results at a predetermined frequency combination, a judgment is made on whether there is abnormal vibration behavior in the wind turbine pitch bearing, thereby realizing the fault identification of the wind turbine pitch bearing.

[0040] In practice, after completing the multi-directional vibration information correlation fusion processing in step S2, fused vibration information reflecting the synchronous correlation characteristics of multi-directional vibration data is obtained. Based on this, step S3 is further used to characterize the cooperative variation relationship between different frequency components in the fused vibration signal, so as to enhance the modulation characteristics caused by fault excitation and suppress the influence of Gaussian random noise.

[0041] For fused vibration information Introducing frequency offset parameter (Used to describe the relative positional relationship between different frequency components, thereby indicating the frequency coordination change caused by the combined effects of structural nonlinearity and fault excitation), construct a multi-frequency joint statistical result to characterize the coordination change characteristics of multiple frequency components, the expression of which is: ; in, This indicates that the fused vibration information is present at the frequency. and frequency offset parameters Multi-frequency joint statistical results under the given conditions Indicates frequency; This represents the frequency offset parameter, used to describe the relative positional relationship between different frequency components.

[0042] The aforementioned multi-frequency joint statistical results are obtained by jointly calculating the fused vibration information at multiple frequency positions. They reflect the cooperative behavior between different frequency components in the fused vibration information and can simultaneously consider the amplitude relationship and phase consistency between different frequency components. When modulation caused by fault excitation exists in the fused vibration signal, the fused vibration information will exhibit cooperative variation characteristics at the dominant frequency and its related offset frequencies, thereby causing the joint statistics to show corresponding... A significant response is generated at the location.

[0043] Since Gaussian random noise lacks stable phase correlation between different frequency components, its contribution to the joint statistics tends to weaken under statistical average. Therefore, this multi-frequency joint statistical processing can effectively suppress noise interference and highlight fault-related characteristics under weak modulation conditions.

[0044] Based on the aforementioned multi-frequency joint statistical results, the response intensity at different frequencies and frequency offset combinations is analyzed. When a significant enhancement occurs at a specific frequency combination, it can be determined that there is an abnormal multi-frequency coordinated change phenomenon in the fused vibration information, and based on this, the possible fault state in the wind turbine pitch bearing can be identified.

[0045] Through the above steps, step S3 achieves an explicit characterization of the fault excitation modulation mechanism without relying on a single frequency feature, providing richer and more robust criterion information for wind turbine pitch bearing fault identification.

[0046] Specifically, in one implementation of this invention, an accelerometer with a sensitivity of 10 mV / g is used to collect vibration signals, and the sensor is installed on the surface of the pitch bearing housing. When collecting fault vibration signals, the pitch bearing containing the faulty outer ring bearing and the sensor are first mounted on a pitch bearing test bench, and then driven by a motor. The test data is then collected using a data acquisition device and transmitted to a computer for subsequent data processing and analysis using MATLAB software. The rotational speed of the shaft containing the bearing is 3 RPM, and the sampling frequency is 38580 Hz. Figure 2 The faulty bearing used in the experiment.

[0047] The acquired signal is subjected to a Fourier transform to obtain its spectrum. Then, envelope demodulation is used to process the signal to obtain its envelope spectrum. Figure 3 The waveform, spectrum, and envelope spectrum of the horizontal vibration signal are shown in the embodiments of the present invention. Figure 4 The waveform, spectrum, and envelope spectrum of the vertical vibration signal in this embodiment of the invention. It is difficult to observe features suitable for fault diagnosis in the waveform, spectrum, and envelope spectrum of signals in both directions.

[0048] Input as follows Figure 3 and Figure 4 The vibration signal shown is used to calculate the synchronous correlation between vibration responses in different directions, and based on the joint changes between different frequency components in the fused vibration information, multi-frequency joint statistical results are generated. The multi-frequency joint statistical results are as follows: Figure 5 As shown.

[0049] Clearly, the multi-frequency joint statistical results show a distinct characteristic frequency of outer ring faults. This result is consistent with actual conditions, thus confirming the effectiveness of this invention for monitoring the condition of wind turbine pitch bearings.

[0050] In summary, this invention provides a method for multi-directional vibration collaborative fault identification of wind turbine pitch bearings: Vibration signals from different directions of the wind turbine pitch bearing are simultaneously acquired at the same measuring point; correlation enhancement processing is performed on the multi-directional vibration signals to extract the consistent response components caused by the same fault excitation; based on this, a multi-frequency joint statistical result is constructed to characterize the collaborative variation characteristics of multiple frequency components; high-order frequency collaborative analysis is performed on the enhanced fused signal to characterize the modulation features and nonlinear frequency coupling characteristics caused by the fault excitation; the operating state of the pitch bearing is judged based on the multi-frequency collaborative characteristics, and the fault identification result is output. This invention fully utilizes the physical correlation between multi-directional vibration data, and improves the distinguishability of weak fault features in a strong noise background through statistical collaborative enhancement and multi-frequency joint characterization. It is suitable for fault identification of wind turbine pitch bearings under varying wind speeds and complex operating conditions, and has strong noise resistance and engineering applicability. Specifically, the following beneficial effects can be achieved by adopting the method of this invention: 1. Overcoming the limitations of unidirectional sensing and improving the observability of fault excitation: This invention constructs a collaborative analysis mechanism for multi-directional vibration data by simultaneously collecting vibration information from multiple structural directions at the same measuring point. This avoids the problem of missing fault information caused by the sensitivity of installation direction in traditional unidirectional measurement, and enables the fault response excited by the complex stress conditions of the pitch bearing to be fully characterized in multiple directions, thereby improving the observability and stability of fault excitation from the source.

[0051] 2. Enhanced weak fault feature analysis capability through multi-layer statistical fusion: By introducing correlation-based frequency fusion processing and multi-frequency joint statistical analysis, this invention strengthens the consistency components in vibration information in different directions, effectively suppresses directional unrelated noise and random interference, and significantly amplifies the weak modulation features hidden in the background of strong noise, thereby achieving in-depth analysis of early and weak fault features.

[0052] 3. Explicitly characterize frequency co-variation to improve the reliability of diagnostic criteria: This invention does not rely on a single frequency amplitude change as a criterion. Instead, it constructs a statistical description of multi-frequency co-variation to jointly characterize the phase consistency and energy coupling characteristics between different frequency components. This transforms the basis for fault identification from local frequency anomalies to multi-frequency co-variations, significantly improving the robustness and anti-misjudgment ability of diagnostic results.

[0053] 4. Adaptable to complex working conditions and possesses good engineering generalization ability: For the complex working conditions of wind turbine pitch bearings such as wind speed fluctuations, frequent pitch loads and strong aerodynamic excitation, this invention effectively reduces the impact of operating condition fluctuations on diagnostic results through multi-directional information collaboration and statistical enhancement mechanisms. It can maintain stable fault identification performance under different speed levels, load conditions and environmental noise levels, and has good engineering adaptability and promotion application value.

[0054] Example 2 Based on the same inventive concept, the present invention also provides a multi-directional vibration collaborative fault identification system for wind turbine pitch bearings, used to implement the multi-directional vibration collaborative fault identification method for wind turbine pitch bearings described in the foregoing embodiments. The system includes: a data acquisition module, a correlation analysis module, and a fault identification module. The data acquisition module is used to time-align the vibration signals obtained at the same structural measuring point along different directions during the operation of the wind turbine pitch bearing, so as to obtain multi-directional vibration data with time consistency. The correlation analysis module is used to calculate the synchronous correlation between vibration responses in different directions in the frequency domain based on the multi-directional vibration data, forming fused vibration information that reflects the common excitation effect; The fault identification module is used to generate multi-frequency joint statistical results based on the joint changes between different frequency components in the fused vibration information, and to judge the operating status of the wind turbine pitch bearing based on the change characteristics of the multi-frequency joint statistical results, thereby realizing fault identification of the wind turbine pitch bearing.

[0055] Furthermore, the data acquisition module performs time-aligned acquisition of vibration signals obtained from the same structural measuring point along different directions during the operation of the wind turbine pitch bearing. The process of obtaining multi-directional vibration data with time consistency includes: Multiple directionally sensitive vibration acquisition units are installed at the same physical location on the pitch bearing of the wind turbine. Based on the vibration acquisition unit, the vibration signal of the wind turbine pitch bearing is acquired in real time and synchronously in different directions to obtain multi-directional vibration data with time consistency.

[0056] Furthermore, the correlation analysis module calculates the synchronous correlation between vibration responses in different directions in the frequency domain based on the multi-directional vibration data, forming fused vibration information reflecting the common excitation effect. This process includes: The multi-directional vibration data are subjected to Fourier transform to obtain a frequency domain representation; Based on the frequency domain representation, the synchronous correlation between vibration responses in different directions is calculated in the frequency domain to obtain fused vibration information that reflects the common excitation effect in multi-directional vibration data.

[0057] Furthermore, the process of calculating the synchronous correlation between vibration responses in different directions in the frequency domain based on the frequency domain representation, and obtaining fused vibration information to reflect the common excitation effect in multi-directional vibration data, includes: ; in, This indicates that the multi-directional vibration data sequence is in frequency The result of the synchronization consistency measurement at the location, i.e., the fused vibration information; and These represent the amplitude coefficients of the vibration response in different directions; superscript Indicates the complex conjugate symbol; This represents the equivalent frequency response of the structure to fault excitation. This represents the equivalent weighting factor introduced by the structural transitivity. Represents the mathematical expectation; and These represent the first and second digits of the multi-directional vibration data sequence, respectively. Next and first The arrival time of the secondary fault impact It represents the imaginary unit.

[0058] Furthermore, the process by which the fault identification module generates multi-frequency joint statistical results based on the joint changes between different frequency components in the fused vibration information includes: By introducing a frequency offset parameter, combined statistical analysis is performed on the fused vibration information at multiple frequency locations to generate multi-frequency joint statistical results.

[0059] Furthermore, by introducing a frequency offset parameter, the process of combining and statistically analyzing the fused vibration information at multiple frequency locations to generate multi-frequency joint statistical results includes: ; in, This indicates that the fused vibration information is present at the frequency. and frequency offset parameters Multi-frequency joint statistical results under certain conditions.

[0060] Furthermore, the fault identification module determines the operating status of the wind turbine pitch bearing based on the changing characteristics of the multi-frequency joint statistical results. The process of fault identification of the wind turbine pitch bearing includes: Based on the variation amplitude of the multi-frequency joint statistical results at a predetermined frequency combination, a judgment is made on whether there is abnormal vibration behavior in the wind turbine pitch bearing, thereby realizing the fault identification of the wind turbine pitch bearing.

[0061] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for identifying a multi-directional vibration collaborative fault of a wind turbine variable pitch bearing, characterized in that, The method comprises: The method comprises: The method comprises: The method comprises:

2. The wind turbine variable pitch bearing multi-directional vibration co-fault identification method of claim 1, wherein, The method comprises: The method comprises: The method comprises:

3. The method of claim 1, wherein, The method comprises: The method comprises: The method comprises:

4. The method of claim 3, wherein, The system comprises: a data acquisition module, an association analysis module, and a fault identification module; ； wherein, represents the synchronization consistency measure result of the multi-directional vibration data sequence at frequency , i.e., the fused vibration information; and respectively represent the amplitude coefficients of the vibration responses in different directions; the superscript represents the complex conjugate symbol; represents the equivalent frequency response of the structure to the fault excitation, represents the equivalent weighting factor introduced by the structure transfer characteristic; represents the mathematical expectation; and respectively represent the arrival time of the th and the th fault impact in the multi-directional vibration data sequence, represents the imaginary unit.

5. The method of claim 4, wherein, The data acquisition module is configured to collect vibration signals obtained in different directions at the same structural measurement point of a wind turbine variable pitch bearing during operation of the wind turbine variable pitch bearing, and obtain multi-directional vibration data with time consistency; The data acquisition module is configured to collect vibration signals obtained in different directions at the same structural measurement point of a wind turbine variable pitch bearing during operation of the wind turbine variable pitch bearing, and obtain multi-directional vibration data with time consistency; 6. The method of claim 5, wherein, ​ ； wherein, represents the joint statistical result of the fusion vibration information in frequency and the frequency offset parameter under the condition.

7. The method of claim 1, wherein, ​ ​ 8. A wind turbine variable pitch bearing multi-directional vibration collaborative fault identification system, the system being configured to implement the wind turbine variable pitch bearing multi-directional vibration collaborative fault identification method of any one of claims 1-7, wherein, ​ ​ The correlation analysis module is configured to calculate, based on the multi-direction vibration data, a synchronous correlation between different direction vibration responses in a frequency domain, and form fused vibration information reflecting a common excitation effect. The fault identification module is configured to generate a multi-frequency joint statistical result according to a joint change between different frequency components in the fused vibration information, and judge the running state of the pitch bearing of the wind driven generator according to a change feature of the multi-frequency joint statistical result, so as to realize fault identification of the pitch bearing of the wind driven generator.

Images