An on-line monitoring device and detection method for the pitch bearing of a wind turbine

Abstract

An on-line monitoring device for a pitch bearing of a wind turbine, comprising biaxial acceleration sensors A, B, and C respectively arranged on the inner surface of the blade root; the biaxial acceleration sensors A, B, and C are connected to a hub collector; the hub collector is connected to a wireless router; the wireless router is connected to a data server; the steps of the on-line monitoring method are as follows: Step 1, connect the monitoring device; Step 2, the hub collector extracts the vibration data of the pitch bearing and the key operating condition data of the unit, and the hub sensor transmits the vibration data of the pitch bearing and the key operating condition data of the unit to the industrial router in the nacelle through a wireless transmission method; the industrial router sends the collected data to the data server; Step 3, the data server analyzes the data and performs diagnostic analysis according to the analysis; it has the characteristics of simple device structure, convenient installation, and can realize on-line monitoring.

Description

Technical Field

[0001] The invention belongs to the technical field of wind turbine condition monitoring, and particularly relates to an on-line monitoring device and a detection method for a pitch bearing of a wind turbine. Background Art

[0002] China's wind power industry is currently in a stage of rapid development, and its role in optimizing the energy structure and promoting energy conservation and emission reduction is becoming increasingly prominent. Areas rich in wind energy are usually areas with very harsh climate conditions, and the alternating loads on the entire fan are very complex.

[0003] As a link connecting the blade and the fan hub, the pitch bearing of the blade is as complex in force as the blade itself. If the whole fan blade is unbalanced, the pitch bearing will be the first to be tested. If there are large cracks in the inner or outer ring of the pitch bearing, the fan blade may directly fall off. Therefore, the damage of the pitch bearing cannot be ignored. On the other hand, the root bolts of the wind turbine blade, as key fasteners connecting the flange of the pitch bearing and the hub flange, and the flange of the pitch bearing and the blade flange, are subjected to complex and variable loads such as irregular vibration, alternating load, and impact load, and are prone to loosening or fracture problems. Local bolt loosening or fracture will cause rapid fatigue of the surrounding bolts, and the fracture of the root bolt is more likely to cause the problem of the pitch gear being jammed and unable to retract the blade, resulting in potential safety hazards.

[0004] Using the ultrasonic scheme to realize the state monitoring of the bolts of the pitch bearing flange, its disadvantages are: the ultrasonic scheme has high cost; the deployment process is complex (slight deviation in the process will lead to unavailable data); the reliability is poor, and the ultrasonic wafer is easy to fail. Using the corner scheme to realize the state monitoring of the bolts of the pitch bearing flange, its disadvantages are: the corner scheme is invalid for the failure mode of bolts with non-rotating nuts and directly cut-off bolts.

[0005] The literature with the patent name (CN201811007469.2) a method for detecting faults of a wind turbine pitch bearing discloses an off-line detection device and method for a pitch bearing. Its measurement method is that the fan is shut down, the impeller is locked, an acceleration sensor is deployed, then sensors and collectors are arranged on the pitch bearing to be measured, and then manual pitch operation is carried out to collect data for analysis, realizing the collection of vibration data of the pitch bearing and the diagnosis of the faults of the pitch bearing body. Its disadvantages are: the pitch state is artificially created, and effective data collection is realized through manual pitching; on-line monitoring cannot be realized, and the real-time diagnosis cannot be guaranteed; only the diagnosis of the faults of the pitch bearing body can be realized, and the diagnosis of the loosening of the root bolts cannot be realized; only single-axis sensors are deployed radially on the pitch bearing, and only the axial vibration signal of the pitch bearing is monitored.

[0006] The patent document with the patent name (CN201711008220.9) is a wind turbine blade health monitoring system based on multi-sensor information, which is used to monitor blade structural damage, surface damage, pitch angle deviation, and icing. That is, the monitoring object is the blade. The involved acceleration sensors are installed on the surface of the cavity at one-third of the blade length from the blade root inside the blade, and the involved strain sensors are installed at the blade root.

[0007] Currently, there is no monitoring system for the pitch bearings (the bearings themselves and flange bolts) of wind turbine units, especially the online monitoring method. There are some systems that can achieve offline monitoring of pitch bearings, and there are also products that can achieve online monitoring of blade root bolts. However, this kind of online monitoring can generally only monitor the inner ring bolts of pitch bearings, and the installation method is very complex, and the system reliability is also relatively poor. Summary of the Invention

[0008] To overcome the deficiencies of the above-mentioned prior art, the purpose of the present invention is to provide an online monitoring device and detection method for the pitch bearings of wind turbine units, which have the characteristics of simple device structure, convenient installation, and can achieve online monitoring.

[0009] To achieve the above purpose, the technical solution adopted by the present invention is:

[0010] An online monitoring device for the pitch bearings of wind turbine units includes biaxial acceleration sensors A, B, and C respectively arranged on the inner surface of the blade root; the biaxial acceleration sensors A, B, and C are connected to the hub collector through shielded signal cables; the hub collector is connected to the wireless router in the nacelle; the wireless router is connected to the data server in the main control room.

[0011] The biaxial acceleration sensors A, B, and C are respectively arranged on the inner surface of the blade root; the biaxial acceleration sensors A, B, and C are respectively 10 - 15 cm away from the inner ring of the pitch bearing.

[0012] The wireless router adopts an industrial router.

[0013] The wireless router is connected to the data server in the main control room through the fan ring network switch.

[0014] An online monitoring method for the pitch bearings of wind turbine units includes the following steps:

[0015] Step 1: Paste the biaxial acceleration sensors A, B, and C on the inner surfaces of the roots of three groups of blades respectively. Before bonding, fix a special base on the biaxial acceleration sensors with bolts; connect the biaxial acceleration sensors A, B, and C to the hub collector through shielded signal cables; then connect the hub collector to the wireless router in the nacelle, and connect the wireless router to the data server in the main control room through the fan ring network switch; connect the data server to the forward isolator; connect the forward isolator to the external network data server.

[0016] Step 2: The hub collector collects the vibration data of the pitch bearing through the biaxial acceleration sensors A, B, and C, synchronously collects the key operating condition parameters of the wind turbine through the PLC or SCADA server, and the hub sensor transmits the vibration data of the pitch bearing and the key operating condition data to the industrial router in the nacelle through wireless transmission; the industrial router sends the collected vibration data and the key operating condition data of the unit to the data server in the main control room through the wind farm ring network;

[0017] Step 3: The data server analyzes the vibration data and the key operating condition data of the unit obtained through the wind farm ring network. If the wind turbine is in the pitch operation state, the state diagnosis and analysis of the pitch bearing body are carried out; if it is in the non-pitch operation state, the loosening diagnosis and analysis of the pitch bearing bolts are carried out.

[0018] For the pitch operation state, the judgment method is as follows: collect the key operating condition data of the unit with a length of T seconds at a frequency of 1 Hz, and then count the deviation between the maximum value and the minimum value of the pitch angle data among them. When the deviation between the maximum and minimum values of the pitch angle reaches 30 degrees, it is determined as the pitch operation state; setting it to 30 degrees can not only accurately judge most of the pitch states, but also avoid excluding the small pitch changes during the normal operation of the wind turbine.

[0019] For the non-pitch operation state, the judgment method is as follows: collect the key operating condition data of the unit with a length of T seconds at a frequency of 1 Hz, and then count the deviation between the maximum value and the minimum value of the pitch angle data among them. When the deviation between the maximum and minimum values of the pitch angle is less than 5 degrees and the active power is greater than 10 kWh at this time, it is judged as the non-pitch operation state.

[0020] The method for diagnosing and analyzing the loosening of the pitch bearing bolts is as follows: Using the axial vibration data of the biaxial acceleration sensors in the non-pitching operation state, perform high-pass filtering on this set of vibration data with a cut-off frequency of 1 Hz, and then calculate the effective value of the filtered vibration data. Make a horizontal comparison of the effective values of the axial vibration data of the biaxial acceleration sensors of the three blade pitch bearings or a vertical comparison with historical data. When it is found that the effective value of the axial vibration data of a certain biaxial acceleration sensor of the pitch bearing is significantly higher than that of the other two pitch bearings, or there is a significant increase compared with historical data, the bolt loosening problem can be detected in a timely manner.

[0021] The method for diagnosing and analyzing the state of the pitch bearing body is as follows: For the radial data of the biaxial acceleration sensors in the pitching state, screen the pitching process data according to the key operating condition data of the wind turbine collected synchronously; then perform high-pass filtering on this set of vibration data with a cut-off frequency of 1 Hz, and finally calculate the kurtosis value of the filtered vibration data; under normal circumstances, the kurtosis value of the radial vibration data of the pitch bearing is below 3, and when its kurtosis value exceeds 3, it is judged that the pitch bearing is damaged.

[0022] The beneficial effects of the present invention are:

[0023] The present invention can realize the loosening diagnosis of the bolts connecting the pitch bearing and the blade and the bolts connecting the pitch bearing and the hub only by installing a biaxial acceleration sensor at the root of each blade, based on the vibration transmission characteristics (even if it is installed on the inner ring of the pitch bearing, the abnormal vibration of the outer ring bolts will be transmitted to the inside through the structure).

[0024] The present invention picks up the axial vibration data of the pitch bearing, extracts energy and other effective characteristic values, and realizes the fastening state diagnosis of the bolts connecting the pitch bearing and the blade and the bolts connecting the pitch bearing and the hub. When the bolts are loose, timely warning can be realized.

[0025] By picking up the radial vibration data of the pitch bearing and the key operating condition data of the unit, the pitching process data is screened, and the effective value and other characteristics of the pitching process data are extracted to realize the damage diagnosis of the inner ring, cage, rolling elements and outer ring of the pitch bearing. When there is an abnormality in the pitch bearing body, timely warning can be given.

[0026] The data server monitoring software can give corresponding warnings and alerts according to the fault levels of the pitch bolts and the pitch bearing body.

[0027] Based on the vibration principle, the present invention is simple in installation and deployment, has a long service life, controllable cost, can realize online monitoring, is convenient for detection, and has outstanding cost performance compared with other monitoring means.

[0028] The system is reliable. The acceleration monitoring solution has been widely experimentally applied in the field of wind turbine condition monitoring. Its reliability is obvious from the sensor to the analog data acquisition unit.

[0029] Online monitoring can be achieved. Acceleration data can be transmitted to the wind farm server in real time. If necessary, all historical data can be fully saved.

[0030] Simple deployment. Only one biaxial acceleration sensor needs to be installed at the root of each blade. Adopting the adhesive method will not cause any harm to the blade. The entire deployment process does not require calibration or zero setting operations.

[0031] Using point measurement to detect the whole, with comprehensive functions. By only installing one biaxial acceleration sensor at the root of each blade, it is possible to achieve looseness diagnosis of the bolts connecting the pitch bearing and the blade and the bolts connecting the pitch bearing and the hub according to the vibration transmission characteristics (even if the abnormal vibration of the outer ring bolts is installed inside the inner ring of the pitch bearing, it will be transmitted to the inside through the structure).

[0032] Controllable cost and great promotion value. Only one biaxial acceleration sensor is installed on each blade, and the installation is simple and reliable, without special maintenance. Therefore, the overall cost is relatively low. Description of the drawings

[0033] Figure 1 It is the principle block diagram of the pitch bearing health status monitoring device of the present invention for wind turbines.

[0034] Figure 2 It is the schematic diagram of the installation positions of biaxial acceleration sensor A, biaxial acceleration sensor B, and biaxial acceleration sensor C of the present invention.

[0035] Figure 3 It is the schematic diagram of the specific installation positions and directions of biaxial acceleration sensor A, biaxial acceleration sensor B, and biaxial acceleration sensor C of the pitch bearing of the present invention.

[0036] Figure 4 It is the flow chart of valid data selection of the present invention.

[0037] Figure 5 It is the time-domain waveform diagram of the axial vibration signal of the pitch bearing for bolt looseness diagnosis.

[0038] Figure 6 It is the time-domain waveform diagram of the radial vibration signal during the pitch process for pitch bearing body fault diagnosis.

[0039] Figure 7 It is the trend diagram of the effective values of the axial accelerations of three pitch bearings.

[0040] Figure 8 It is the time-domain waveform diagram of the radial signal of a certain pitch bearing during the pitch process.

[0041] Figure 9 This is a flowchart of the pitch bearing bolt monitoring method of the present invention.

[0042] Figure 10 This is a flowchart of the pitch bearing body monitoring method of the present invention.

[0043] In the figure: 1 - Biaxial acceleration sensor A, 2 - Biaxial acceleration sensor B, 3 - Biaxial acceleration sensor C, 4 - Data acquisition unit, 5 - Inner surface of the blade, 6 - Leading edge of the blade, 7 - Trailing edge of the blade, 8 - Blade, 9 - Outer ring of the pitch bearing, 10 - Spherical hub cover, 11 - Cylindrical hub wall, 12 - Inner ring of the pitch bearing, 13 - Gear on the inner ring. Specific embodiments

[0044] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0045] An on-line monitoring device for the pitch bearing of a wind turbine includes biaxial acceleration sensors A, B, and C respectively arranged on the inner surface of the blade root; the biaxial acceleration sensors A, B, and C are connected to the hub collector through shielded signal cables; the hub collector is connected to a wireless router in the nacelle; the wireless router is connected to the data server in the main control room.

[0046] The biaxial acceleration sensors A, B, and C are respectively arranged on the inner surface of the blade root (the combined mold surface of the leading edges of the three groups of blades); the biaxial acceleration sensors A, B, and C are respectively 10 - 15 cm away from the inner ring of the pitch bearing. The position selection method is as follows: The installer stands in the nacelle facing the hub and locks the blade to be installed to the three-point horizontal direction. Enter the hub. The installer faces the locked blade, and the biaxial acceleration sensor is installed at the horizontal position of the left blade root, that is, the combined mold surface of the leading edge. The biaxial acceleration sensor is 10 - 15 cm away from the inner ring of the pitch bearing. The first axis of the biaxial acceleration sensor points to or away from the blade tip, and the second axis points to or away from the installation surface. The wire outlet direction of the sensor can be upward or downward.

[0047] The wireless router used is an industrial router.

[0048] The wireless router is connected to the data server in the main control room through the fan ring network switch.

[0049] The working principle of the present invention is:

[0050] Since the hub collector, the wind turbine PLC, and the SCADA server are within the same local area network, the hub collector synchronously collects the key operating condition data of the wind turbine while collecting the vibration data of the pitch bearing. The hub collector transmits the vibration data and the key operating condition data of the unit to the industrial router in the nacelle through wireless or wired transmission; the industrial router is connected to the data server in the main control room through the wind farm ring network, and the industrial router sends the collected data to the data server in the main control room through the wind farm ring network; the data server is connected to the forward isolator, and the forward isolator is connected to the external network data server. The data server can push the complete internal network monitoring module to the external network data server through the forward isolator through the push software module for remote diagnosticians to view.

[0051] See Figure 1 , Figure 1 is a health status monitoring system for the pitch bearing of a wind turbine, including biaxial acceleration sensors A1, B2, and C3 respectively installed on the inner surface of the blade root. The biaxial acceleration sensors collect the vibration data of the pitch bearing; the biaxial acceleration sensors are connected to the hub collector through shielded signal cables; the hub collector synchronously extracts the key operating condition data of the wind turbine through the PLC or SCADA system; the hub collector is connected to the wireless router in the nacelle, and the hub sensor transmits the vibration data of the pitch bearing and the key operating condition data of the unit to the industrial router in the nacelle through wireless transmission; the industrial router is connected to the data server in the main control room through the wind farm ring network, and the industrial router sends the collected vibration data and operating condition data to the data server in the main control room through the wind farm ring network; the data server is connected to the forward isolator, and the forward isolator is connected to the external network data server. The data server can push the internal network monitoring data to the external network data server through the forward isolator through the push software for remote diagnosticians to view.

[0052] Figure 2 The black squares in

[0052] are the installation positions of the biaxial acceleration sensors A1, B2, and C3 respectively, that is, the leading edge position of the blade root. Figure 2 In Figure 2 , the data acquisition unit 4 is installed in the reserved hole position inside the hub in the form of a wall-mounted pendant. The data acquisition unit 4 has three or more M12 aviation connectors, and the electrical interface of the biaxial acceleration sensor is also an M12 aviation connector, which can be connected by plugging in finished cables, which is convenient and reliable. Due to the existence of the pitch mechanism, the cable cannot directly run from inside the blade to the hub. Instead, a telescopic cable bridge needs to be built between the hub and the blade using springs, etc., and then the cable is fixed to the spring in an S shape. The data acquisition unit can synchronously collect a total of six vibration signals from three biaxial acceleration sensors (i.e., biaxial acceleration sensors A1, B2, and C3).

[0053] Figure 3 Taking the pitch structure type driven by the inner ring (the internal gear ring of the pitch bearing) as an example, the specific installation positions of the sensors are expressed. That is, the dual-axis acceleration sensor 1 is installed at the mold clamping position of the blade leading edge 6 and the inner surface 5 of the blade. The position of the dual-axis acceleration sensor A1 deviates from the blade end face 10 - 15 cm of the inner ring 12 of the pitch bearing. To ensure accurate and consistent installation positions, the following positioning method can be adopted: The installer stands in the nacelle facing the hub and locks the blade to be installed to the three-point horizontal direction. Enter the hub. The installer faces the locked blade 8, and the dual-axis acceleration sensor is installed at the horizontal position of the blade root on the installer's left hand, that is, the mold clamping surface of the blade leading edge 6. The first axis of the dual-axis acceleration sensor points to the axial direction of the pitch bearing (towards the blade tip or away from the blade tip), and the second axis points to the radial direction of the pitch bearing (towards or away from the sensor plane). The dual-axis acceleration sensor is bonded with polyurethane glue. Before bonding, a special base is fixed to the bottom of the sensor with bolts, so that the glue will not cover the sensor during bonding, which is convenient for later maintenance.

[0054] A pitch bearing is provided at the upper end of the cylindrical hub wall 11 and is connected to the blade 8 through the pitch bearing; the blade 8 includes a blade leading edge 6 and a blade trailing edge 7; the pitch bearing includes a pitch bearing outer ring 9 and a pitch bearing inner ring 12; an inner ring upper gear 13 is provided inside the inner ring 12; a spherical hub cover covers the outside of the cylindrical hub wall 11.

[0055] The hub acquisition unit can synchronously acquire acceleration data and key operating condition data of the unit. After the acceleration data and key operating condition data of the unit are transmitted to the data server, effective data screening work can be carried out. In this example, the acceleration data sampling rate is 1280 Hz and the duration is 120 seconds. The sampling rate of the operating condition signal is 1 Hz and the duration is 120 seconds. As Figure 4 shown, first judge whether the wind turbine is in the operating state according to the key operating condition data of the unit. When the power data in the operating condition data > 10 kW, it is in operation; ≤ 0 kW, it is in shutdown; if it is shutdown data, this group of data is not specially processed and belongs to invalid data; if it is operating data, then pitch judgment is carried out; process the pitch angle data in the 120-second operating condition data collected, and calculate the deviation between the maximum value and the minimum value in this group of data. If the maximum and minimum deviation of the pitch angle < 5 degrees, it is in the non-pitching state; if the maximum and minimum deviation of the pitch angle > 30 degrees, it is in the pitching state; if it is in the non-pitching state, then select the axial vibration data for bolt loosening diagnosis and analysis, as Figure 5 shown in the waveform; if it is in the pitching state, then select the radial vibration data for pitch bearing body fault diagnosis and analysis, as Figure 6 shown in the waveform.

[0056] Figure 7It is a trend graph drawn by extracting the effective values of the axial acceleration data of each group during non-pitch operation of the three pitch bearings. Pitch bearing one, pitch bearing two, and pitch bearing three are monitored by biaxial acceleration sensors A1, B2, and C3 respectively. It can be seen from the graph that the effective value of the axial acceleration data of pitch bearing three is significantly higher, indicating that there may be problems such as loosening or fracture of the bolts of this pitch bearing.

[0057] Figure 8 The time-domain waveform after filtering the vibration signal during the pitch process of a certain abnormal pitch bearing. It can be seen from the graph that when there is an abnormality in the pitch bearing body, abnormal impacts will appear in the signal. By calculating its kurtosis value, it is found that it is much greater than 3. Therefore, the kurtosis value can well diagnose the faults of the pitch bearing body.

[0058] The purpose of the present invention is to provide a condition monitoring system for the pitch bearings of wind turbines based on vibration information and operating conditions, aiming to continuously monitor the conditions of the pitch bearing body and flange bolts of the wind turbine for 24 hours, and to evaluate the health status of the pitch bearings in real time and efficiently.

[0059] The types of the pitch bearings described include two types. One is that the inner ring is connected to the blade and the outer ring is connected to the hub; the other is that the outer ring is connected to the blade and the inner ring is connected to the hub. The pitch methods include pitch by gear ring meshing and pitch by belt.

[0060] The flange bolts include the bolts connecting the pitch bearing and the blade, and also include the bolts connecting the pitch bearing and the hub.

[0061] Since the biaxial acceleration sensor of the present invention is pasted on the inner surface of the blade root, and a transfer base is installed at the bottom of the sensor before bonding, the adhesive will not cover the sensor, which is convenient for later maintenance. The type of adhesive is polyurethane structural adhesive, and its specific position setting standard is as follows: The installer stands in the nacelle facing the hub and locks the blade to be installed to the three-point horizontal direction. Enter the hub. The installer faces the locked blade and installs the biaxial acceleration sensor at the horizontal position of the left blade root, that is, the front edge mold surface. The distance between the biaxial acceleration sensor and the inner ring of the pitch bearing is 10-15 cm. The first axis of the biaxial acceleration sensor points to or away from the blade tip, and the second axis points to or away from the installation surface. The wire outlet direction of the sensor can be upward or downward.

[0062] The hub collector has at least 6 IEPE signal channels, and every two IEPE signal channels are integrated on an aviation connector, which is convenient for quick plug-in wiring of the three blades. The hub collector has the function of wired network communication or wireless network communication and can adapt to wired or wireless communication scenarios.

[0063] The key operating condition data of the unit can be obtained from the fan PLC or from the SCADA system in the main control room. The key operating condition data of the unit at least includes: wind speed, active power, ambient temperature, pitch angle of pitch bearing 1, pitch angle of pitch bearing 2, pitch angle of pitch bearing 3. The synchronous acquisition refers to starting the acquisition of operating conditions and acceleration signals at the same time.

[0064] The industrial wireless router in the nacelle has a wireless WIFI communication function and can receive and transmit stable wireless WIFI signals.

[0065] The industrial router is directly connected to the wind farm ring network through an Ethernet cable.

[0066] See Figures 9 - 10 , an on-line monitoring method for the pitch bearing of a wind turbine, comprising the following steps:

[0067] Step 1, paste the biaxial acceleration sensor A, biaxial acceleration sensor B, and biaxial acceleration sensor C on the inner surfaces of the roots of the three groups of blades respectively. Before bonding, first fix a special base at the bottom of the sensor with bolts; connect the biaxial acceleration sensor A, biaxial acceleration sensor B, and biaxial acceleration sensor C to the hub collector through a shielded signal cable; then connect the hub collector to the wireless router in the nacelle, and connect the wireless router to the data server in the main control room through the fan ring network switch; connect the data server to the forward isolator; connect the forward isolator to the external network data server.

[0068] Step 2, the hub collector collects the vibration data of the pitch bearing through the biaxial acceleration sensor A, biaxial acceleration sensor B, and biaxial acceleration sensor C, synchronously collects the key operating condition parameters of the wind turbine through the PLC or SCADA server, and the hub sensor transmits the vibration data of the pitch bearing and the key operating condition data of the unit to the industrial router in the nacelle through a wireless transmission method; the industrial router sends the collected vibration data and the key operating condition data of the unit to the data server in the main control room through the wind farm ring network;

[0069] Step 3, the data server analyzes the vibration data and the key operating condition data of the unit obtained through the wind farm ring network. If the wind turbine is in the pitch operation state, the state diagnosis and analysis of the pitch bearing body are carried out; if it is in the non-pitch operation state, the diagnosis and analysis of the loosening of the pitch bearing bolts are carried out.

[0070] For the pitch operation state, the judgment method is as follows: collect the key operating condition data of the unit with a length of T seconds at a frequency of 1 Hz, and then count the deviation between the maximum value and the minimum value of the pitch angle data among them. When the deviation between the maximum and minimum values of the pitch angle reaches 30 degrees, it is determined as the pitch operation state; setting it to 30 degrees can not only accurately judge most of the pitch states, but also avoid excluding small pitch changes during the normal operation of the wind turbine.

[0071] For the described non-pitch operation state, the judgment method is as follows: Collect the key operating condition data of the unit with a length of T seconds at a frequency of 1 Hz, and then statistically analyze the deviation between the maximum and minimum values of the pitch angle data among them. When the deviation between the maximum and minimum values of the pitch angle is less than 5 degrees and the active power is greater than 10 kWh at this time, it is judged as the non-pitch operation state.

[0072] For the diagnosis and analysis of the loosening of the pitch bearing bolts, the method is as follows: Utilize the axial vibration data of the dual-axis acceleration sensor in the non-pitch operation state, perform high-pass filtering on this set of data with a cut-off frequency of 1 Hz, and then obtain the effective value of the filtered data. Conduct a horizontal comparison of the effective values of the axial data of the pitch bearings of the three blades or a vertical comparison with historical data. When it is found that the effective value of a certain pitch bearing is significantly higher than that of the other two pitch bearings, or there is a significant increase compared with historical data, the problem of bolt loosening can be detected in a timely manner.

[0073] For the diagnosis and analysis of the state of the pitch bearing body, the method is as follows: Intercept the radial data of the sensor in the pitch state, and only intercept the pitch process data according to the simultaneously collected operating condition data; then perform high-pass filtering on this set of data with a cut-off frequency of 1 Hz, and finally obtain the kurtosis value of the filtered data; under normal circumstances, the kurtosis value of the pitch bearing is below 3. When its kurtosis value exceeds 3, it is judged that the pitch bearing is damaged.

Claims

1. An on-line monitoring method for the pitch bearing of a wind turbine, characterized in that, It includes the following steps: Step 1: Paste biaxial acceleration sensors A, B, and C on the surface of the mold closing of the leading edges of three groups of blades respectively. Before bonding, fix a special base on the biaxial acceleration sensors with bolts; connect biaxial acceleration sensors A, B, and C to the hub collector through shielded signal cables; then connect the hub collector to the wireless router in the nacelle, and connect the wireless router to the data server in the main control room through the fan ring network switch; connect the data server to the forward isolator; connect the forward isolator to the external network data server; Step 2: The hub collector collects the vibration data of the pitch bearing through biaxial acceleration sensors A, B, and C, synchronously collects the key operating condition data of the wind turbine through the PLC or SCADA server, and the hub sensor transmits the vibration data of the pitch bearing and the key operating condition data of the unit to the industrial router in the nacelle through wireless transmission; the industrial router sends the collected vibration data and the key operating condition data of the unit to the data server in the main control room through the wind farm ring network; Step 3: The data server analyzes the vibration data and the key operating condition data of the unit obtained through the wind farm ring network. If the wind turbine is in the pitch operation state, the pitch bearing body state diagnosis and analysis are carried out according to the radial vibration data of the biaxial acceleration sensor; if it is in the non-pitch operation state, the pitch bearing bolt loosening diagnosis and analysis are carried out according to the axial vibration data of the biaxial acceleration sensor.

2. The on-line monitoring method for the pitch bearing of a wind turbine unit according to claim 1, characterized in that For the pitch operation state mentioned above, the judgment method is: collect the key operating condition data of the unit with a length of T seconds at a frequency of 1 Hz, and then count the deviation between the maximum value and the minimum value of the pitch angle data. When the deviation between the maximum and minimum values of the pitch angle reaches 30 degrees, it is determined as the pitch operation state.

3. The on-line monitoring method for the pitch bearing of a wind turbine unit according to claim 1, characterized in that, For the non-pitch operation state mentioned above, the judgment method is: collect the key operating condition data of the unit with a length of T seconds at a frequency of 1 Hz, and then count the deviation between the maximum value and the minimum value of the pitch angle data. When the deviation between the maximum and minimum values of the pitch angle is less than 5 degrees and the active power is greater than 10 kWh at this time, it is judged as the non-pitch operation state.

4. The on-line monitoring method for the pitch bearing of a wind turbine unit according to claim 1, characterized in that, For the pitch bearing bolt loosening diagnosis and analysis mentioned above, the method is: use the axial vibration data of the biaxial acceleration sensor in the non-pitch operation state, perform high-pass filtering on this group of vibration data with a cut-off frequency of 1 Hz, then obtain the effective value of the filtered vibration data, and make a horizontal comparison of the effective values of the axial vibration data of the biaxial acceleration sensors of the pitch bearings of the three blades or a vertical comparison with historical data. When it is found that the effective value of a certain pitch bearing is significantly higher than that of the other two pitch bearings, or there is a significant increase compared with historical data, the bolt loosening problem can be detected in time.

5. A method for online monitoring of a pitch bearing of a wind turbine unit according to claim 1, characterized in that, For the pitch bearing body state diagnosis and analysis mentioned above, the method is: Screen the radial data of the dual-axis acceleration sensor in the pitch state, and only screen the pitch process data according to the key operating condition data of the unit collected synchronously; then perform high-pass filtering on this set of vibration data with a cut-off frequency of 1 Hz, and finally calculate the kurtosis value of the filtered vibration data; under normal circumstances, the kurtosis value of the pitch bearing is below 3, and when its kurtosis value exceeds 3, it is judged that the pitch bearing is damaged.

Images