Wind driven generator fault early warning method and system based on optical fiber acoustic emission technology

Abstract

The invention discloses a wind driven generator fault early warning method and system based on an optical fiber acoustic emission technology, and relates to the field of generator fault monitoring. The invention aims to monitor the operation state of the wind driven generator in real time through an optical fiber acoustic emission technology so as to find mechanical faults, abnormal vibration or structural damage in time and ensure the safe and stable operation of the wind driven generator. The system comprises a plurality of optical fiber acoustic emission sensors which are arranged at key parts of the wind driven generator and are used for acquiring acoustic emission signals in real time. The signal is transmitted to the data processing unit through the optical fiber, and the fault of the wind driven generator is diagnosed and positioned. By analyzing acoustic emission signals, early warning is performed in advance, and further deterioration of faults or shutdown of equipment is avoided. The fault early warning system for the wind driven generator has the advantages of high sensitivity, strong anti-interference capability, adaptability to severe environments and the like, and is suitable for on-line health monitoring and preventive maintenance of the wind driven generator.

Description

Technical Field

[0001] The invention relates to the field of generator fault monitoring, and in particular to a wind generator fault early warning method and system based on optical fiber acoustic emission technology. Background Art

[0002] Traditional wind turbine fault diagnosis methods mostly rely on regular inspections and offline detection. This method is not only inefficient, but also difficult to capture the instantaneous information of the fault, making it difficult to accurately locate and predict the fault. In recent years, with the rapid development of sensor technology and signal processing technology, online monitoring technologies based on parameters such as vibration, temperature, and current have gradually been applied to wind turbine fault diagnosis. However, these technologies can often only monitor the macroscopic performance after the fault occurs, and there are still certain limitations for early warning and precise positioning of faults. In particular, the accuracy of traditional sensor monitoring is not high, which can easily cause inaccurate wind turbine fault warnings, leading to false alarms or missed alarms, thereby causing accidents. Summary of the invention

[0003] The purpose of the present invention is to overcome one or more deficiencies of the prior art and to provide a wind turbine fault warning method and system based on optical fiber acoustic emission technology.

[0004] The objective of the present invention is achieved through the following technical solutions:

[0005] A wind turbine fault early warning method based on optical fiber acoustic emission technology, comprising the following steps:

[0006] S1. Install optical fiber acoustic emission sensors in the wind turbines, and connect the optical fiber acoustic emission sensors of all wind turbines to the underground optical fiber network through optical fibers;

[0007] S2. The optical fiber signal is connected to the multi-channel acoustic emission collection and processing system after being polled by the optical fiber demodulator;

[0008] S3. Analyze and process the received optical fiber signal to identify the fault status of the wind turbine to be monitored;

[0009] S4. Provide real-time fault warning of fault status through local area network or cloud platform.

[0010] Furthermore, in step S2, the optical fiber signal is connected to the same channel and polled in time sequence, so that the single channel can be expanded into multiple channels.

[0011] Furthermore, in step S2, after passing through the polling fiber optic demodulator, the optical signal containing wavelength and phase data in the optical fiber is converted into an electrical signal representing the fault vibration of the wind turbine, and the electrical signal is further connected to the multi-channel acoustic emission acquisition and processing system through a coaxial cable.

[0012] Furthermore, in step S3, the processing process of the multi-channel acoustic emission collection and processing system specifically includes: processing and analyzing the signals of the optical fiber acoustic emission sensors of all wind turbines, analyzing the signal characteristics, and comparing them with the existing fault models in the multi-channel acoustic emission collection and processing system to identify the fault status of each monitored wind turbine.

[0013] A wind turbine fault warning system based on optical fiber acoustic emission technology, comprising:

[0014] Fiber optic acoustic emission sensor system, fiber optic signal transmission system, polling fiber optic demodulation system and multi-channel acoustic emission acquisition and processing system;

[0015] Fiber optic acoustic emission sensor system: A fiber optic acoustic emission sensor is installed in each wind turbine to transmit fiber optic signals;

[0016] Optical fiber signal transmission system: connected to the optical fiber acoustic emission sensor system, specifically the optical fiber acoustic emission sensors installed in all wind turbines, connected via optical fiber to transmit optical fiber signals;

[0017] Polling fiber demodulation system: connected to the fiber signal transmission system, used for polling fiber demodulator to poll the fiber signal;

[0018] Multi-channel acoustic emission acquisition and processing system: connected to the polling fiber optic demodulation system via a coaxial cable, used to analyze and warn the fiber optic signal.

[0019] Through the multi-channel acoustic emission acquisition and processing system, the signals of the optical fiber acoustic emission sensors of all wind turbines are processed and analyzed, the signal characteristics are analyzed, and at the same time, they are compared with the existing fault models in the multi-channel acoustic emission acquisition and processing system software to identify the fault status of each monitored wind turbine. The fault status of each wind turbine monitored in the wind farm is connected to the local area network of the wind farm or to the cloud service platform for real-time fault warning, including:

[0020] The fault status of each wind turbine monitored in the wind farm can be connected to the local area network of the wind farm and displayed in real time on the large screen in the wind farm monitoring room or the main control room. The operator can monitor it or connect to the cloud service platform to remotely provide real-time fault warning.

[0021] The real-time fault warning status is divided into normal status, fault repair status and fault shutdown status, and the fault level is set, and the wind farm operation and maintenance team makes operation and maintenance decisions.

[0022] Furthermore, the polling type optical fiber demodulation system includes an optical fiber demodulator and a polling input device.

[0023] The beneficial effects of the present invention are:

[0024] (1) Use distributed fiber optic acoustic emission sensors to monitor the fault status of wind turbines in real time, thereby performing real-time health diagnosis of wind turbines, and providing early warning and precise positioning of key wind turbine components;

[0025] (2) Each wind turbine is equipped with only optical fiber sensors and signal transmission optical fibers. The sensors and signal transmission optical fibers are made of the same optical fiber, which is small in size and light in weight, and is suitable for the installation and arrangement of sensors for key components of wind turbines.

[0026] (3) Since the optical signal loss inside the optical fiber is basically zero, the optical signal can be transmitted for hundreds or thousands of kilometers without distortion. Therefore, the optical signal collected by the optical fiber acoustic emission sensor can be directly transmitted through tens of kilometers of optical fiber to a distant optical fiber demodulator and collection equipment, achieving long-distance transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is a schematic diagram of the structure of a wind turbine fault warning system based on fiber-optic acoustic emission technology;

[0028] Figure 2 This is an example diagram of polling of 64-channel input to 8-channel output of a polling fiber optic demodulator. DETAILED DESCRIPTION

[0029] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0030] Example 1

[0031] See also Figure 1 , provides a wind turbine fault warning system based on fiber optic acoustic emission technology, including:

[0032] Fiber optic acoustic emission sensor system, fiber optic signal transmission system, polling fiber optic demodulation system and multi-channel acoustic emission acquisition and processing system;

[0033] Fiber optic acoustic emission sensor system: A fiber optic acoustic emission sensor is installed in each wind turbine to transmit fiber optic signals;

[0034] Optical fiber signal transmission system: connected to the optical fiber acoustic emission sensor system, specifically the optical fiber acoustic emission sensors installed in all wind turbines, connected via optical fiber to transmit optical fiber signals;

[0035] Polling fiber demodulation system: connected to the fiber signal transmission system, used for polling fiber demodulator to poll the fiber signal;

[0036] Multi-channel acoustic emission acquisition and processing system: connected to the polling fiber optic demodulation system via a coaxial cable, used to analyze and warn the fiber optic signal.

[0037] Through the multi-channel acoustic emission acquisition and processing system, the signals of the optical fiber acoustic emission sensors of all wind turbines are processed and analyzed, the signal characteristics are analyzed, and at the same time, they are compared with the existing fault models in the multi-channel acoustic emission acquisition and processing system software to identify the fault status of each monitored wind turbine. The fault status of each wind turbine monitored in the wind farm is connected to the local area network of the wind farm or to the cloud service platform for real-time fault warning, including:

[0038] The fault status of each wind turbine monitored in the wind farm can be connected to the local area network of the wind farm and displayed in real time on the large screen in the wind farm monitoring room or the main control room. The operator can monitor it or connect to the cloud service platform to remotely provide real-time fault warning.

[0039] The real-time fault warning status is divided into normal status, fault repair status and fault shutdown status, and the fault level is set, and the wind farm operation and maintenance team makes operation and maintenance decisions.

[0040] According to the properties of wind turbines and wind field characteristics, several fiber optic acoustic emission sensors are arranged in key components of wind turbines, including but not limited to rotating blades, variable pitch bearings, motors, main shafts, towers, etc. The number of sensors is not limited. In the actual sensor arrangement plan, considering the economic cost and installation efficiency, one fiber optic acoustic emission sensor is installed on a single rotating blade, one fiber optic acoustic emission sensor is installed on the variable pitch bearing, one fiber optic acoustic emission sensor is installed on the main shaft, and one fiber optic acoustic emission sensor is installed on the tower. A total of eight fiber optic acoustic emission sensors can monitor all key components of the entire wind turbine.

[0041] Since the wind farm covers a range of less than 100 kilometers in diameter, and optical fiber transmission signals can ensure that the optical signal within this distance is not distorted, and the basic loss is 0, all optical fiber acoustic emission sensors can be connected to the underground optical fiber ring network built by the wind farm through optical fibers of several kilometers to tens of kilometers. Optical fiber itself is waterproof and does not carry electricity, so optical fiber acoustic emission sensors can be directly installed on outdoor wind turbines without considering lightning protection.

[0042] In the actual installation process, due to the large number of fiber optic cables, each fiber optic cable is individually numbered and all the fibers of a single wind turbine need to be gathered together. The fiber optic cables are numbered xx#xx. For example, 2#2 represents the second fiber optic sensor of the second wind turbine. After this arrangement, all the fibers can be neatly and uniformly connected to the underground fiber optic ring network of the wind farm.

[0043] See also Figure 2 , consider monitoring 8 wind turbines in a wind farm at the same time, and install 8 acoustic emission sensors on each wind turbine. The acoustic emission sensors at the same position of all wind turbines are connected to the input ports of the same group of polling optical fiber demodulators, that is, the optical fiber acoustic emission sensors numbered 1#1, 2#1, 3#1, 4#1, 5#1, 6#1, 7#1, and 8#1 are connected to the I1 group input ports of the optical fiber modulator, representing the signal input of the same position 1 of different wind turbines. Similarly, the 8 input ports of the I1 group are uniformly output from the O1 output port, and the signals of 1#1, 2#1, 3#1, 4#1, 5#1, 6#1, 7#1, and 8#1 are input in sequence at a time interval of 1s, and then all output from the O1 output port.

[0044] After the operation of the polling fiber optic demodulator, 8 channels are required for acoustic emission collection and processing, which corresponds to the number of acoustic emission sensors installed on a single wind turbine.

[0045] The fault models of different parts of wind turbines are different, especially the fault diagnosis model of composite materials such as blades is more complicated. There are two common acoustic emission signal analysis methods, one is the classic parameter analysis method, and the other is the wave analysis method. The parameter analysis method is a simple and commonly used acoustic emission signal analysis method. The parameter analysis method performs a preliminary transformation on the obtained acoustic emission signal and analyzes and processes the acoustic emission source through the acoustic emission parameters defined physically or mathematically. Commonly used characteristic parameters include ring count, energy, rise time, pulse duration, average signal level and effective voltage (RMS), etc. Wave analysis methods include spectrum analysis methods and wavelet analysis methods. Fault pattern recognition also includes cluster analysis and currently popular artificial intelligence recognition algorithms. Through data analysis, the fault diagnosis models of different components are determined and input into a wind turbine fault real-time early warning system software based on optical fiber acoustic emission technology provided by the present invention, and the data is analyzed in real time to perform fault warning.

[0046] The real-time fault warning status is divided into normal status, fault repair status and fault shutdown status, and the fault level is set, and the wind farm operation and maintenance team makes operation and maintenance decisions. Local area network: A central monitoring server is set up in the local area network to receive and process fault data from each wind turbine. Configure the corresponding database and monitoring software to realize the storage, analysis and visualization of fault data. The monitoring software performs real-time analysis on the received fault data and immediately triggers a warning notification once an abnormality or potential fault is detected. The warning notification can be sent to relevant personnel through broadcast, email, SMS and other methods within the local area network so that timely measures can be taken to troubleshoot and repair the fault. When accessing the cloud service platform: Connect the system to the cloud service platform through the Internet or a dedicated network. Create corresponding accounts and projects on the cloud service platform to receive and manage fault data from each wind turbine. Configure the monitoring and warning functions of the cloud platform, including setting warning rules, notification methods, etc. The cloud platform performs real-time analysis on the received fault data and uses data mining, machine learning and other technologies to extract potential fault characteristics and abnormal patterns. Once a possible fault is detected, the cloud platform will automatically trigger an early warning notification and send it to relevant personnel in real time through SMS, email, phone, etc. The cloud platform can also provide intelligent management functions such as fault prediction, performance evaluation, and maintenance suggestions to help operation and maintenance personnel better understand the equipment status and formulate reasonable maintenance plans. Whether it is connected to the wind farm LAN or the cloud service platform, real-time monitoring of the fault status of wind turbines can be achieved to improve operation and maintenance efficiency.

[0047] The fiber optic acoustic emission sensor of this embodiment specifically includes: the fiber optic acoustic emission sensor includes two types: a fiber optic ring acoustic emission sensor and a fiber optic grating acoustic emission sensor, has waterproof and lightning protection functions, withstands a temperature range of -273°C to 1000°C, measures a vibration frequency range including but not limited to 20k to 250kHz, and has a sensitivity greater than 100dB.

[0048] The present embodiment discloses a method and system for early warning of wind turbine faults based on optical fiber acoustic emission technology, which uses distributed optical fiber acoustic emission sensors to monitor the fault status of wind turbines in real time, thereby performing real-time health diagnosis of wind turbines, and is expected to achieve early fault warning and precise positioning of key components of wind turbines. Each wind turbine is only equipped with an optical fiber sensor and a signal transmission optical fiber. The sensor and the signal transmission optical fiber are the same optical fiber, which is small in size and light in weight, and is very suitable for the installation and arrangement of sensors of key components of wind turbines. Since the loss of optical signals inside the optical fiber is basically zero, the optical signal can be transmitted for hundreds or thousands of kilometers without distortion. Therefore, the optical signal collected by the optical fiber acoustic emission sensor can be directly transmitted to a distant optical fiber demodulator and collection equipment through tens of kilometers of optical fiber. Therefore, by using a real-time early warning system for wind turbine faults based on optical fiber acoustic emission technology disclosed by the present invention, it is only necessary to build a set of optical fiber demodulators and acoustic emission collection and processing systems in the main control room of the wind farm to synchronously monitor all wind turbines in the wind farm.

[0049] Example 2

[0050] A wind turbine fault warning method based on optical fiber acoustic emission technology is provided, and the specific steps are as follows:

[0051] S1. Install optical fiber acoustic emission sensors in the wind turbines, and connect the optical fiber acoustic emission sensors of all wind turbines to the underground optical fiber network through optical fibers;

[0052] S2. The optical fiber signal is connected to the multi-channel acoustic emission collection and processing system after being polled by the optical fiber demodulator;

[0053] S3. Analyze and process the received optical fiber signal to identify the fault status of the wind turbine to be monitored;

[0054] S4. Provide real-time fault warning of fault status through local area network or cloud platform.

[0055] The specific implementation is as follows:

[0056] 1. Overview of wind farm:

[0057] A large coastal wind farm has 50 wind turbines with a single unit capacity of 3MW, distributed in a vast coastal tidal flat area. The area is rich in wind resources, but the environment is relatively harsh, with high humidity, high salinity and frequent strong winds.

[0058] 2. System Installation and Initialization

[0059] Fiber optic acoustic emission sensor arrangement (step S1):

[0060] In the nacelle of each wind turbine, fiber optic acoustic emission sensors are installed for key components such as gearboxes, main shaft bearings, generators, etc. For example, a sensor is installed near the high-speed shaft and low-speed shaft of the gearbox to fully capture abnormal sound signals during gear meshing; at the main shaft bearing, three sensors are evenly arranged along the circumference of the bearing to ensure that the vibration and acoustic emission conditions at different positions of the bearing can be accurately monitored. These sensors are connected by fiber optic cables that are resistant to high temperatures, corrosion, and have good mechanical strength. The optical fibers of all wind turbines are finally converged and connected to the underground optical fiber network. The underground optical fiber network uses armored optical cables to enhance its ability to resist external damage, and is laid along the pre-laid cable trenches between the wind turbine arrays to ensure the stability and reliability of signal transmission.

[0061] Signal transmission and demodulation (step S2):

[0062] The optical fiber signal transmitted from the wind turbine first reaches the polling optical fiber demodulator located near the wind farm control center. The polling optical fiber demodulator polls each optical fiber signal in turn at intervals of 5 milliseconds according to the set time sequence. For example, it first receives the optical fiber signal of wind turbine No. 1, performs demodulation processing, converts the wavelength and phase information in the optical signal into the corresponding electrical signal, and then switches to the signal processing of wind turbine No. 2, and so on. The processed electrical signal is connected to the multi-channel acoustic emission collection and processing system through a coaxial cable. In this process, the polling optical fiber demodulator has automatic calibration and signal gain adjustment functions to adapt to the signal changes caused by factors such as distance and signal strength differences of different wind turbines, ensuring that each signal can be accurately converted into an electrical signal for subsequent system analysis.

[0063] 3. Signal analysis and fault identification (step S3):

[0064] Data preprocessing: After receiving the electrical signal, the multi-channel acoustic emission collection and processing system first filters the signal to remove high-frequency noise and low-frequency interference signals. For example, a bandpass filter is used to set the passband range to 10kHz-100kHz. This frequency band can effectively retain the characteristic acoustic emission signal generated when the key components of the wind turbine fail. Then the filtered signal is amplified so that the signal amplitude meets the requirements of subsequent analysis and processing.

[0065] Feature extraction and analysis: Extract features from the preprocessed signal, calculate the signal's time domain features such as root mean square value (RMS), peak factor, kurtosis, etc., and frequency domain features such as spectrum peak frequency, frequency center of gravity, etc. For example, when a gearbox has a wear failure, the frequency domain features of its acoustic emission signal will show energy concentration within a specific frequency range. By monitoring and analyzing the spectrum peak frequency, the operating status of the gearbox can be preliminarily determined. At the same time, calculate the signal's time-frequency features, and use methods such as short-time Fourier transform (STFT) or wavelet transform to obtain the signal's energy distribution at different times and frequencies, so as to have a more comprehensive understanding of the signal's characteristics.

[0066] Fault identification and judgment: The extracted characteristic parameters are compared with the pre-established fault model library in the system. The fault model library is obtained by analyzing a large amount of historical fault data and training the machine learning algorithm. For example, the fault model of the main shaft bearing includes the characteristic parameter ranges under various conditions such as normal state, inner ring wear, outer ring wear, and rolling element failure. When the degree of match between the monitored signal characteristics and a certain fault model exceeds the set threshold (such as 80%), it can be determined that the corresponding component of the fan is faulty and the fault type can be determined.

[0067] 4. Fault warning and processing (step S4):

[0068] Generation and release of early warning information: Once the fault status of the wind turbine is identified, the multi-channel acoustic emission collection and processing system immediately generates fault early warning information. The early warning information includes the wind turbine number, faulty component, fault type, fault occurrence time, and fault severity assessment. This information is transmitted to the monitoring center of the wind farm through the local area network, and the operation and maintenance personnel are prompted with eye-catching pop-up windows and alarm sounds on the large screen of the monitoring center. At the same time, the system can also send the early warning information to the remote operation and maintenance center through the cloud platform so that professional technicians can obtain information in a timely manner and provide remote technical support.

[0069] Operation and maintenance response and processing: After receiving the early warning information, the operation and maintenance personnel will formulate a corresponding maintenance plan based on the severity and type of the fault. For minor faults, such as early minor wear of bearings, they can be arranged to be handled during the next scheduled maintenance; for serious faults, such as broken teeth in the gearbox, a repair team will be immediately organized to go to the site for emergency repairs. During the maintenance process, maintenance personnel can refer to the detailed information provided by the fault warning system to inspect and repair the faulty parts in a targeted manner to improve maintenance efficiency and accuracy. After the maintenance is completed, the faulty fan is retested and monitored to ensure that it resumes normal operation, and the maintenance record is fed back to the fault warning system so that the system can update and optimize the fault model.

[0070] The above is only a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the form disclosed herein, and should not be regarded as excluding other embodiments, but can be used in various other combinations, modifications and environments, and can be modified within the scope of the concept described herein through the above teachings or the technology or knowledge of the relevant field. The changes and modifications made by those skilled in the art shall not deviate from the spirit and scope of the present invention, and shall be within the scope of protection of the claims attached to the present invention.

Claims

1. A wind turbine fault warning method based on optical fiber acoustic emission technology, characterized in that the steps include: S1. Connect the fiber optic acoustic emission sensors of all wind turbines to the underground fiber optic network through optical fibers; S2. The optical fiber signal is connected to the multi-channel acoustic emission collection and processing system after being polled by the optical fiber demodulator; S3. Analyze and process the received optical fiber signal to identify the fault status of the wind turbine to be monitored; S4. Provide real-time fault warning of fault status through local area network or cloud platform.

2. A wind turbine fault early warning method based on optical fiber acoustic emission technology according to claim 1, characterized in that: In step S2, the optical fiber signal is connected to the same channel and polled in time sequence, so that the single channel can be expanded into multiple channels.

3. A wind turbine fault early warning method based on optical fiber acoustic emission technology according to claim 1, characterized in that: In step S2, after passing through the polling fiber demodulator, the optical signal containing wavelength and phase data in the optical fiber is converted into an electrical signal representing the fault vibration of the wind turbine generator, and the electrical signal is further connected to the multi-channel acoustic emission acquisition and processing system through a coaxial cable.

4. A wind turbine fault early warning method based on optical fiber acoustic emission technology according to claim 1, characterized in that: In step S3, the processing process of the multi-channel acoustic emission collection and processing system specifically includes: processing and analyzing the signals of the optical fiber acoustic emission sensors of all wind turbines, analyzing the signal characteristics, and comparing them with the existing fault models in the multi-channel acoustic emission collection and processing system to identify the fault status of each monitored wind turbine.

5. A wind turbine fault warning system based on optical fiber acoustic emission technology, characterized in that: include: Fiber optic acoustic emission sensor system, fiber optic signal transmission system, polling fiber optic demodulation system and multi-channel acoustic emission acquisition and processing system; Fiber optic acoustic emission sensor system: A fiber optic acoustic emission sensor is installed in each wind turbine to transmit fiber optic signals; Optical fiber signal transmission system: connected to the optical fiber acoustic emission sensor system, specifically the optical fiber acoustic emission sensors installed in all wind turbines, connected via optical fiber to transmit optical fiber signals; Polling fiber demodulation system: connected to the fiber signal transmission system, used for polling fiber demodulator to poll the fiber signal; Multi-channel acoustic emission acquisition and processing system: connected to the polling fiber optic demodulation system via a coaxial cable, used to analyze and warn the fiber optic signal.

6. A wind turbine fault warning system based on fiber optic acoustic emission technology according to claim 5, characterized in that: The polling type optical fiber demodulation system comprises an optical fiber demodulator and a polling input device.

Images