Operation monitoring method and device for wind turbine bearing holder

Abstract

An operation monitoring method for a wind turbine bearing holder, which method belongs to the technical field of data supervision and prediction, and is used for solving the technical problem of it not being conducive to performing rapid maintenance of an abnormal wind turbine generator set due to the fact that it is difficult to perform real-time online supervision and prediction on the operation situation of an existing wind turbine generator set, and that there is also certain hysteresis in the fault reporting of the wind turbine generator set. The method comprises: by means of a plurality of sensing chips which are preset in a wind turbine bearing holder, performing multi-point multi-cluster-head assisted tracking on the operation state of the wind turbine bearing holder, so as to obtain point tracking information (S101); performing motion vector correction and prediction related to points on the point tracking information, which has been subjected to orthogonal coverage, so as to obtain a circumferential motion trajectory of the points (S102); performing filtering processing regarding irregular trajectories on the current circumferential motion trajectory, so as to obtain an ideal circumferential motion trajectory (S104); then, performing positioning comparison of probabilistic centroids on an ideal circumferential spatial region and a predicted circumferential spatial region (S105); and obtaining operation monitoring information of a wind turbine bearing (S106). Further provided is an operation monitoring device (500) for a wind turbine bearing holder.

Description

A method and device for monitoring the operation of a wind turbine bearing retainer Technical Field

[0001] The present application relates to the field of data monitoring and prediction, and in particular to an operation monitoring method and device for a wind turbine bearing retainer. Background Art

[0002] As a green energy source, wind power has been developing at an unexpectedly rapid pace since the 1980s. With the increasing maturity of related technologies, it has become a widely used renewable energy power generation method. Wind turbines operate under complex alternating loads for a long time, placing increasingly stringent requirements on wind power transmission systems. Bearings, as key components of wind turbines, play a vital role in ensuring the reliability of the entire unit.

[0003] Currently, the failure rate of gearboxes and motors in wind turbines is high, and most of these failures are caused by faulty wind turbine bearings. Existing methods generally use vibration sensors installed inside wind turbines to monitor the operating status of the wind turbine main shaft, detecting abnormalities in voltage, current, power, and other signals within the wind turbine, and indirectly monitoring the operation of the wind turbine bearings or bearing retainers.

[0004] However, the aforementioned vibration sensor monitoring method is difficult to monitor the specific operating characteristics of wind turbine bearings or wind turbine bearing retainers. It has a certain lag, and detection is often only possible after a fault occurs and severe vibration occurs. Furthermore, wind turbine generator systems are often located in remote areas, leading to long maintenance cycles, which can easily cause irreparable damage to wind turbines, blades, hubs, and other equipment, resulting in a significant waste of resources. Furthermore, it is difficult to conduct real-time online monitoring and prediction of the wind turbine main shaft's operating status, and thus cannot provide timely and early warning feedback to maintenance personnel, increasing the operating and maintenance costs of wind turbines.

[0005] Summary of the Invention

[0006] The embodiments of the present application provide an operation monitoring method and device for a wind turbine bearing retainer, which is used to solve the following technical problems: the operation status of existing wind turbines is difficult to monitor and predict online in real time, and there is a certain lag in reporting faults of wind turbines, which is not conducive to the rapid maintenance of abnormal wind turbines and increases a lot of maintenance costs.

[0007] The embodiments of this application adopt the following technical solutions:

[0008] On the one hand, an embodiment of the present application provides an operation monitoring method for a wind turbine bearing retainer, comprising: performing multi-point multi-cluster head assisted tracking of the operation status of the wind turbine bearing retainer through a plurality of sensor chips preset in the wind turbine bearing retainer to obtain point tracking information; wherein, the point tracking information includes: target coordinate data and corresponding acceleration data of all sensor chips in any time period; performing orthogonal coverage of the point tracking information; and performing motion vector correction and prediction on the point tracking information after orthogonal coverage to obtain the circular motion trajectory of the point; wherein, the circular motion trajectory includes: current circular motion trajectory and predicted circular motion trajectory; according to the vibration acceleration of the wind turbine bearing retainer, obtaining the point tracking information. The instantaneous vibration circular trajectory of the wind turbine bearing retainer; through the instantaneous vibration circular trajectory, the current circular motion trajectory in the circular motion trajectory is filtered out of irregular trajectories to obtain the ideal circular motion trajectory of the wind turbine bearing retainer; the corresponding ideal circular motion trajectory and the predicted circular motion trajectory are respectively generated into corresponding ideal circular space areas and corresponding predicted circular space areas; and the probability centroids of the ideal circular space areas and the predicted circular space areas are positioned and compared to obtain the predicted overlapping space area; according to the predicted overlapping space area, whether there is any abnormality in the operation of the wind turbine bearing retainer is judged, and the operation monitoring information of the wind turbine bearing is obtained to complete the operation monitoring of the wind turbine group.

[0009] The beneficial effect of this application is that by monitoring the running trajectory of the wind turbine bearing retainer in the wind turbine main bearing, it can monitor the abnormal operation of the wind turbine bearing retainer in real time based on the error comparison between the predicted overlapping space area and the actual overlapping space area, and then predict whether there are abnormalities in the wind turbine main shaft and the wind turbine group, which is conducive to timely feedback of possible fault problems to maintenance personnel, so that they can arrive at the scene as soon as possible to prevent the further fermentation of potential faults. It is conducive to real-time online supervision and prediction of the operation status of the wind turbine group, reducing the delay time for reporting potential faults of the wind turbine group, facilitating the rapid maintenance of wind turbine groups with abnormalities, reducing maintenance costs, and ensuring the normal power generation efficiency of the wind turbine group.

[0010] In a feasible embodiment, by presetting a plurality of sensor chips in a wind turbine bearing retainer, the operating status of the wind turbine bearing retainer is tracked with the assistance of multiple cluster heads at multiple points to obtain point tracking information, which specifically includes: using a signal acquisition device in a wind turbine generator set to acquire signals from a plurality of sensor chips in the wind turbine bearing retainer to determine whether the wind turbine bearing retainer is operating; wherein the plurality of sensor chips are embedded and evenly distributed in the wind turbine bearing retainer and make the wind turbine bearing retainer meet the rotational balance; when the wind turbine bearing retainer is operating, the signal sending node of the first sensor chip in operation is determined as the main cluster head node, and the signal sending nodes of the second sensor chip and the third sensor chip are determined as adjacent cluster head nodes; wherein , the second sensor chip and the third sensor chip are respectively located at the left and right adjacent positions of the first sensor chip; according to the preset RSSI algorithm and based on the preset time period, the three-dimensional spatial distance calculation of the received signal carrier power is performed on the main cluster head node and the adjacent cluster head node, and the main coordinate data and the adjacent coordinate data in the current time period are obtained respectively; wherein, the main coordinate data and the adjacent coordinate data are both three-dimensional coordinate data; according to the least squares algorithm, the adjacent coordinate data in the current time period are processed with the minimum value of the average spatial distance, and the spatial distance after the minimum value processing is calculated with the median value to obtain the auxiliary coordinate data; based on the auxiliary coordinate data and the target coordinate data of the first sensor chip, the point tracking information of all sensor chips is determined.

[0011] The beneficial effect of the present application is that by installing multiple sensor chips in the wind turbine bearing retainer, the operation status of the double-row spherical roller bearing can be monitored in real time, that is, the operation posture and operation trajectory of the wind turbine bearing retainer in three-dimensional space can be tracked in real time, and according to the coordinated fusion of multiple cluster heads, based on the adjacent chips of each sensor chip, real-time tracking of each sensor chip is achieved to obtain the main coordinate data and adjacent coordinate data in the current time period.

[0012] In a feasible implementation, based on the auxiliary coordinate data and the target coordinate data of the first sensor chip, the point tracking information of all sensor chips is determined, specifically including: based on the current time period, the auxiliary coordinate data in the adjacent cluster head node and the main coordinate data in the main cluster head node are weightedly fused to obtain the target coordinate data of the first sensor chip; according to the preset WSN grid structure, the signal sending node of the first sensor chip is determined as the adjacent cluster head node, and the target coordinate data is determined as the adjacent coordinate data; by comparing the coordinate data of the fourth sensor chip with the coordinate data of the first sensor chip, the target coordinate data is determined as the adjacent coordinate data. After the target coordinate data of the chip are processed to minimize the average spatial distance, the target coordinate data of the second sensor chip are obtained based on the weighted fusion of the coordinate data; wherein, the fourth sensor chip and the first sensor chip are both placed in adjacent positions on the left and right of the second sensor chip; and so on, the weighted fusion of the relevant coordinate data of all sensor chips is performed to determine the target coordinate data of all sensor chips respectively; and the acceleration data corresponding to the multiple sensor chips in the current time period are obtained; based on the target coordinate data and the corresponding acceleration data of all sensor chips, the point tracking information of all sensor chips is determined.

[0013] The beneficial effect of the present application is that the adjacent cluster head nodes are integrated into the main cluster head node, and the weight fusion of the auxiliary coordinates and the main coordinates is realized, which is conducive to the target tracking of the main cluster head node based on the adjacent auxiliary cluster head nodes, that is, the tracking of the first sensor chip is realized, and finally the coordinate data and corresponding acceleration data of the first sensor chip are obtained. Similarly, the point tracking information of the second, third, fourth and so on sensor chips are obtained respectively, and the point tracking of all sensor chips in the wind turbine bearing retainer is realized.

[0014] In a feasible embodiment, the point tracking information is orthogonally covered, specifically including: based on the orthogonal coverage mechanism, sampling the target coordinate data corresponding to each sensor chip in the point tracking information in the current time period to obtain position data of multiple points related to the target coordinate data; dividing a sampling space area corresponding to the point sampling according to the position data of the multiple points; calibrating the position data of the multiple points with respect to signal strength through the sampling space area to obtain signal strength serial numbers related to the multiple points; judging the motion trend of the position data of the multiple points according to the signal strength serial numbers and the point density in the sampling space area, and determining the point motion trend data based on the current time period based on the point with the largest signal strength in the signal strength serial numbers as the reference point; obtaining the acceleration data corresponding to each sensor chip in the point tracking information; and according to the point motion trend data, making a one-to-one correspondence between the acceleration data and the position data of the multiple points in the sampling space area to generate the current circular motion trajectory based on the point tracking information.

[0015] The beneficial effect of the present application is that the point tracking information of each sensor chip is used to determine the sampling space area in the current time period. Based on the different distances between each sensor core in the sampling space area and the signal acquisition device, different signal strength numbers are formed. This is beneficial for better judging the movement trends of multiple points in the current sampling space area based on the signal strength number and the point density and point position of the sampling space, and then combining it with the acceleration of each point to finally form the current circular motion trajectory.

[0016] In a feasible implementation, the point tracking information after orthogonal coverage is used to perform motion vector correction and prediction on the relevant points to obtain the circular motion trajectory of the points, specifically including: obtaining the point tracking information after orthogonal coverage in the current time period; according to the Lagrange interpolation function and based on the acceleration data in the current circular motion trajectory, performing coordinate position vector prediction of the target coordinate data in the current circular motion trajectory in the next time period to obtain predicted target coordinate data; according to the positioning distance between each point in the current circular motion trajectory, performing acceleration vector correction and prediction on the acceleration data in the current circular motion trajectory in the next time period to obtain predicted target coordinate data; The method comprises the following steps: performing a prediction on the predicted acceleration data; sampling the predicted points of the predicted target coordinate data; dividing the predicted sampling space area corresponding to the predicted target coordinate data; determining the predicted motion trend of multiple points in the predicted sampling space area according to the signal strength sequence numbers of the predicted points in the predicted sampling space area and the corresponding predicted point density; and generating a predicted circular motion trajectory based on the next time period according to the predicted points in the predicted sampling space area and the corresponding predicted acceleration data; and obtaining the circular motion trajectory of the point based on the predicted circular motion trajectory and the current circular motion trajectory.

[0017] The beneficial effect of the present application is that through the Lagrange difference function, the coordinate data and acceleration data of all points in the next time period can be better predicted, and then the prediction and circular motion trajectory corresponding to the current circular motion trajectory can be generated, which is conducive to the subsequent comparative prediction of the operating trajectory of the wind turbine bearing retainer.

[0018] In a feasible embodiment, the instantaneous vibration circular trajectory of the wind turbine bearing retainer is obtained according to the vibration acceleration of the wind turbine bearing retainer, specifically comprising: obtaining the vibration acceleration in the current time period through the vibration sensor in the wind turbine generator set; performing quaternion differential division on the motion trend data in the current circular motion trajectory through the quaternion parameter algorithm to obtain the operating attitude matrix related to the quaternion differential; dividing the vibration acceleration in the current time period into components of each axis in the three-dimensional space according to the operating attitude matrix to obtain vibration vector coordinate data; performing transient fitting of the vibration acceleration and the vibration vector coordinate data on a circular curve to obtain a transient fitting curve; matching the transient fitting curve to the corresponding position based on the three-dimensional space where the wind turbine bearing retainer is located to determine the instantaneous vibration circular trajectory based on the current time period.

[0019] The beneficial effect of the present application is that the vibration acceleration obtained is divided into each axial direction by the vibration sensor preset in the wind turbine, and then the vibration vector data is transiently fitted with a curve to obtain the instantaneous vibration circular trajectory of the vibration acceleration, thereby accurately identifying the offset caused by the impeller load vibration, and obtaining the instantaneous vibration circumferential trajectory of the wind turbine bearing retainer based on the offset state.

[0020] In a feasible embodiment, the instantaneous vibration circular trajectory is used to filter out irregular trajectories of the current circular motion trajectory in the circular motion trajectory to obtain the ideal circular motion trajectory of the wind turbine bearing retainer, specifically including: linearly normalizing the instantaneous vibration circular trajectory and the current circular motion trajectory to obtain an instantaneous vibration circular curve and a current circular curve respectively; wherein the instantaneous vibration circular curve and the current circular curve are both spiral circular curves; performing difference processing on the corresponding coordinate points of the instantaneous vibration circular curve and the current circular curve to obtain the distance of multiple coordinate points; and performing median processing on the distance of the multiple coordinate points to obtain the vibration difference distance; performing curve correction on the current circular curve according to the vibration difference distance to obtain a corrected circular curve; and performing vector processing on the corrected circular curve according to the acceleration data in the current circular motion trajectory to determine the corrected circular motion trajectory; by filtering out irregular trajectories of the current circular motion trajectory within a preset error range using the corrected circular motion trajectory, the ideal circular motion trajectory of the wind turbine bearing retainer is obtained.

[0021] The beneficial effect of the present application is that the error of the current circular motion trajectory is corrected by identifying the instantaneous vibration circular trajectory, and the current circular curve is corrected in time according to the vibration difference distance, so that the actual operating conditions inside the wind turbine can be accurately obtained, that is, the actual circular motion trajectory of the wind turbine bearing retainer.

[0022] In a feasible implementation, corresponding ideal circular space areas and corresponding predicted circular space areas are generated for the ideal circular motion trajectory and the predicted circular motion trajectory, respectively; and the positioning and comparison of the probability centroids of the ideal circular space areas and the predicted circular space areas are performed to obtain a predicted overlapping space area, specifically comprising: generating, based on the ideal circular motion trajectory and the predicted circular motion trajectory, a first spiral cylinder corresponding to the ideal circular space area and a second spiral cylinder corresponding to the predicted circular space area, respectively; wherein the first spiral cylinder and the second spiral cylinder both contain a plurality of point position information; and obtaining the first point position information in the first spiral cylinder; According to the significance of the probability distribution function, the point distribution plane area corresponding to the first point position information is obtained; and the center of mass of the spiral cylinder is located in the point distribution plane area through the probability density function to obtain the first center of mass position information of the first spiral cylinder; the center of mass of the relevant point distribution plane area of ​​the second spiral cylinder is located to obtain the second center of mass position information of the second spiral cylinder; according to the first center of mass position information and the second center of mass position information, a three-dimensional spatial overlap comparison is performed between the first spiral cylinder and the second spiral cylinder in the same time domain and the same space domain to determine the predicted overlap space area that overlaps with the first spiral cylinder and the second spiral cylinder.

[0023] The beneficial effect of the present application is that based on the circular motion trajectory in three-dimensional space in each time period, a motion trajectory quantity with a time component, that is, a spiral cylinder, is constructed, and then according to the center of mass position, the spatial areas where the spiral cylinders are located in different time periods are overlapped and compared to obtain the overlap area under the predicted situation, which is beneficial to the operation monitoring of the wind turbine bearing retainer and the prediction of the overlap of spatial areas in different time periods under normal circumstances.

[0024] In a feasible embodiment, based on the predicted overlapping spatial area, the operation status of the wind turbine bearing retainer is judged to determine whether there is any abnormality, and the operation monitoring information of the wind turbine bearing is obtained, specifically including: according to the actual circular motion trajectory corresponding to the next time period, determining the third spiral cylinder of the actual circumferential spatial area; wherein the actual circular motion trajectory is the point circular motion trajectory of the next time period based on the ideal circular motion trajectory; positioning the center of mass of the third spiral cylinder in the relevant point distribution plane area to obtain the third center of mass position information of the third spiral cylinder; based on the third center of mass position information, performing a three-dimensional spatial overlap comparison in the same time domain and the same spatial domain between the first spiral cylinder and the third spiral cylinder to obtain a real overlapping spatial area; through the spatial area size judgment information of the real overlapping spatial area and the predicted overlapping spatial area, judging whether there is any abnormality in the operation status of the wind turbine bearing retainer, and obtaining the operation monitoring information of the wind turbine bearing, so as to complete the operation monitoring of the wind turbine group.

[0025] On the other hand, an embodiment of the present application also provides an operation monitoring device for a wind turbine bearing retainer, the device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, so that the at least one processor can execute an operation monitoring method for a wind turbine bearing retainer described in any of the above embodiments.

[0026] The beneficial effect of this application is that by monitoring the running trajectory of the wind turbine bearing retainer in the wind turbine main bearing, it can monitor the abnormal operation of the wind turbine bearing retainer in real time based on the error comparison between the predicted overlapping space area and the actual overlapping space area, and then predict whether there are abnormalities in the wind turbine main shaft and the wind turbine group, which is conducive to timely feedback of possible fault problems to maintenance personnel, so that they can arrive at the scene as soon as possible to prevent the further fermentation of potential faults. It is conducive to real-time online supervision and prediction of the operation status of the wind turbine group, reducing the delay time for reporting potential faults of the wind turbine group, facilitating the rapid maintenance of wind turbine groups with abnormalities, reducing maintenance costs, and ensuring the normal power generation efficiency of the wind turbine group. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments described in the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0028] FIG1 is a flow chart of an operation monitoring method for a wind turbine bearing retainer provided in an embodiment of the present application;

[0029] FIG2 is a schematic structural diagram of a double-row spherical roller bearing provided in an embodiment of the present application;

[0030] FIG3 is a schematic diagram of a structure for monitoring the operation of a wind turbine bearing according to an embodiment of the present application;

[0031] FIG4 is a schematic diagram of the distribution of sensor chips of a wind turbine bearing retainer provided in an embodiment of the present application;

[0032] FIG5 is a schematic structural diagram of an operation monitoring device for a wind turbine bearing retainer provided in an embodiment of the present application. DETAILED DESCRIPTION

[0033] In order to enable those skilled in the art to better understand the technical solutions in this application, the following will clearly and completely describe the technical solutions in the embodiments of this application in conjunction with the drawings in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0034] The embodiment of the present application provides a method for monitoring the operation of a wind turbine bearing retainer. As shown in FIG1 , the method for monitoring the operation of a wind turbine bearing retainer specifically includes steps S101 to S106:

[0035] It should be noted that the wind turbine transmission main shaft of a wind turbine in a wind turbine generator system has two wind turbine bearings, generally using double-row spherical roller bearings. Alternatively, for example, FIG2 is a schematic structural diagram of a double-row spherical roller bearing provided in an embodiment of the present application. As shown in FIG2 , the spherical roller bearing has double rows of balls and is fixed and retained by a bearing retainer. During the operation of the wind turbine, the main shaft will bear a certain load pressure, causing the center of mass of the main shaft to deflect. Therefore, the characteristics of the spherical roller bearing can enable the main shaft to achieve a certain offset, and the wind turbine bearing retainer will also perform a certain movement deflection with the offset of the main shaft, that is, not only horizontal rotation, but also offset in various angular directions. Therefore, monitoring the motion trajectory of the wind turbine bearing retainer can achieve operational monitoring of the wind turbine bearing, thereby better reflecting the operational status of the wind turbine main shaft and, therefore, the operational status of the entire wind turbine.

[0036] S101: Using multiple sensor chips pre-installed in the wind turbine bearing retainer, perform multi-point multi-cluster head-assisted tracking of the wind turbine bearing retainer's operating status to obtain point tracking information. The point tracking information includes target coordinate data and corresponding acceleration data of all sensor chips in any time period.

[0037] Specifically, a signal acquisition device in the wind turbine generator system collects signals from multiple sensor chips in the wind turbine bearing holder to determine whether the wind turbine bearing holder is operating. The multiple sensor chips are embedded in the wind turbine bearing holder and evenly distributed, ensuring that the wind turbine bearing holder is rotationally balanced. When the wind turbine bearing holder is operating, the signal transmission node of the operating first sensor chip is determined as the primary cluster head node, and the signal transmission nodes of the second and third sensor chips are determined as adjacent cluster head nodes. The second and third sensor chips are located adjacent to the first sensor chip, respectively.

[0038] In one embodiment, FIG4 is a schematic diagram of the sensor chip distribution of a wind turbine bearing retainer provided in an embodiment of the present application. As shown in FIG4 , when preparing a wind turbine bearing retainer that conforms to a double-row spherical roller bearing, a plurality of sensor chips are placed into the wind turbine bearing retainer through a stamping process or an integral manufacturing process and evenly distributed within the retainer to ensure that the retainer meets dynamic balancing requirements during operation and ensures normal and stable operation of the retainer. Furthermore, according to the arrangement method shown in FIG4 , eight sensor chips are arranged and placed one at a time on the left and right sides. At the same time, the sensor chip has a signal transmission function that transmits real-time coordinate data and acceleration data, so that the signal acquisition device can collect the transmitted signal.

[0039] In one embodiment, Figure 3 is a simplified schematic diagram of a structure for monitoring the operation of wind turbine bearings provided by an embodiment of the present application. As shown in Figure 3, a signal server connected to a signal acquisition device in a wind turbine generator set performs multi-point, multi-cluster head-assisted tracking of the operating wind turbine bearing holder. Cluster heads of different levels are used to perform different tasks at different stages, improving the positioning and tracking accuracy of the sensor chip and reducing the loss rate of target points. Cluster heads are divided into primary cluster head nodes and auxiliary cluster head nodes.

[0040] Furthermore, based on a preset RSSI (Received Signal Strength Indicator) algorithm and within a preset time period, a three-dimensional spatial distance calculation is performed on the primary cluster head node and the adjacent cluster head nodes, which is related to the received signal carrier power. This calculates the primary coordinate data and adjacent coordinate data for the current time period. Both the primary coordinate data and the adjacent coordinate data are three-dimensional coordinate data.

[0041] In one embodiment, the signal receiving device calculates the relative distance of the received signal carrier power between the determined main cluster head node and the adjacent cluster head node based on the RSSI algorithm. as well as Among them, α, β, γ are coordinate reference intermediate quantities, P0 is the standard signal carrier power, P y 、P x 、P z are the carrier powers in the y, x, and z axes respectively, d0 is the preset standard signal receiving distance, and then the main coordinate data and adjacent coordinate data in the current time period are determined respectively.

[0042] Furthermore, according to the least squares algorithm, the adjacent coordinate data in the current time period are processed with the minimum value of the average spatial distance, and the spatial distance after the minimum value processing is calculated with the median value to obtain the auxiliary coordinate data.

[0043] In one embodiment, two adjacent coordinate data within the current time period are represented by a matrix of variable positioning, and matrix A related to the second sensor chip and matrix B corresponding to the third sensor chip are obtained respectively. Because there may be ranging errors, a random error vector is added and combined with a least squares algorithm to perform minimum processing on the average spatial distance between matrix A and matrix B. The average spatial distance after the minimum processing is then median-derivatived, and finally auxiliary coordinate data of the mutual mapping of the two adjacent coordinates is obtained.

[0044] Furthermore, based on the current time period, the auxiliary coordinate data in the adjacent cluster head nodes and the main coordinate data in the main cluster head node are fused with the weights of the relevant coordinate data to obtain the target coordinate data of the first sensor chip.

[0045] As a feasible implementation method, the auxiliary coordinate data in the adjacent cluster head nodes are sent to the main cluster head node, and then the auxiliary coordinate data and the main coordinate data are de-weighted according to the different weight ratios of the auxiliary coordinate data, that is, the auxiliary coordinate data is integrated into the main coordinate data to reduce the tracking and positioning error caused by the single tracking and positioning of the first sensor chip, thereby ensuring the accuracy of the main coordinate data positioning and tracking in the main cluster head node.

[0046] Furthermore, by analogy, according to the preset WSN (wireless sensor network) grid structure, the signal sending node of the first sensor chip is first determined as the adjacent cluster head node, and the target coordinate data is determined as the adjacent coordinate data. After the coordinate data of the fourth sensor chip and the target coordinate data of the first sensor chip are processed to minimize the average spatial distance, and based on the weighted fusion of the coordinate data, the target coordinate data of the second sensor chip is obtained. Among them, the fourth sensor chip and the first sensor chip are both placed in adjacent positions on the left and right of the second sensor chip. By analogy, the weighted fusion of the relevant coordinate data of all sensor chips is performed to determine the target coordinate data of all sensor chips. And the acceleration data corresponding to multiple sensor chips in the current time period is obtained. Then, based on the target coordinate data and corresponding acceleration data of all sensor chips, the point tracking information of all sensor chips is finally determined.

[0047] As a feasible implementation, as shown in Figure 4, the position tracking of the remaining sensors is performed in sequence according to the position tracking of the first sensor chip. Similarly, the second, third, fourth, and so on sensor chips are successively determined as the main cluster head nodes, and the corresponding adjacent sensor chips are successively determined as adjacent cluster head nodes. Ultimately, the target coordinate data of each sensor chip is obtained. Then, combined with the time period in which each sensor chip is located, the acceleration data of each sensor chip in the corresponding time period is obtained. Ultimately, the acceleration data and target coordinate data of each sensor chip are determined as the point tracking information of the point where the sensor chip is located.

[0048] S102: Perform orthogonal overlay on the point tracking information. Perform motion vector correction and prediction on the point tracking information after orthogonal overlay to obtain a circular motion trajectory of the point. The circular motion trajectory includes a current circular motion trajectory and a predicted circular motion trajectory.

[0049] Specifically, based on the orthogonal coverage mechanism, the target coordinate data corresponding to each sensor chip in the point tracking information is sampled in the current time period to obtain the position data of multiple points related to the target coordinate data. Based on the multi-point position data, the sampling space area corresponding to the sampling point is then divided.

[0050] In one embodiment, by extracting any two sampling samples from the point tracking information, the three-dimensional coordinates corresponding to the two points are sampled and divided, and then the two points are orthogonally covered according to the orthogonal mechanism, that is, sample point 1 (X1, Y1, Z1) and sample point 2 (X2, Y2, Z2), and then according to X fin =βX1+αX2,Y fin =βY1+αY2, Z fin=βZ1+αZ2, and obtain the position data of multiple points, that is, the multi-point position coordinates (X fin , Y fin , Z fin ), and then divide the sampling space area corresponding to the sampling point according to the position data extracted by these samples, where α and β are orthogonal covering intermediate quantities.

[0051] Furthermore, the position data of multiple points is calibrated for signal strength using the sampling spatial area to obtain signal strength serial numbers for the multiple points. Based on the signal strength serial numbers and the point density in the sampling spatial area, the movement trend of the position data of the multiple points is determined. The point with the highest signal strength in the signal strength serial number is used as the reference point to determine the point movement trend data based on the current time period.

[0052] In one embodiment, according to The position data of the sampling area is judged on the signal strength, and the signal strength number P(N) of each point is obtained, where L is the point density, μ is the sampling space area, d1 is the signal receiving distance between the point and the signal acquisition device, d0 is the inherent error distance, and ε is the signal chip frequency parameter. Then, according to the signal strength number of each point, each point is marked accordingly. Then, according to the marked number, the movement trend of the position data of multiple points is judged, and the point with the largest signal strength is used as the reference point. Finally, the point movement trend data based on the current time period is determined.

[0053] Furthermore, the acceleration data corresponding to each sensor chip in the point tracking information is first obtained. Based on the point motion trend data, the acceleration data is then matched one-to-one with the position data of multiple points in the sampling space area to generate the current circular motion trajectory based on the point tracking information.

[0054] Furthermore, the point tracking information after orthogonal coverage in the current time period is obtained. According to the Lagrange interpolation function and based on the acceleration data in the current circular motion trajectory, the coordinate position vector of the target coordinate data in the current circular motion trajectory is predicted for the next time period to obtain the predicted target coordinate data.

[0055] In one embodiment, the WSN (wireless sensor network) point coordinates in the point tracking information after orthogonal coverage are all one-time coordinates. The WSN point coordinates in the next time period may enter a new coverage area as they move. Since the mobile wireless sensor network nodes are in a low-speed state and the coverage radius is generally not less than 10m, the point tracking information within the time period of 1s can be within the orthogonal coverage area of ​​the sampling space area in the time period. Then assume that the coordinates corresponding to the k time period are (X k , Y k , X k ), the Lagrangian interpolation function L(k) performs value prediction for the k time period. That is, the Lagrangian interpolation function L(k) performs Lagrangian prediction by obtaining the target coordinate data in the current k time period, so as to obtain the predicted target coordinate data in the k+1 time period (X k+1 , Y k+1 , Z k+1 ).

[0056] As a feasible implementation method, according to Get the Lagrange interpolation function L(i) in the current k time period, where t k is the amount of time in k time period, t k-1 is the amount of time in k-1 time period, L is the point density. Then according to the x-axis coordinate value of k time period: X k =L(t)=L1X k-1 +L2X k-2 , L(t) is the Lagrange interpolation function under the time quantity t, which can further derive the coordinate value X of the x-axis under the time period k+1 k+1 By analogy, we can obtain the y-axis coordinate and z-axis coordinate in the k+1 time period respectively, and finally obtain the predicted target coordinate data (X k+1 , Y k+1 , Z k+1 ).

[0057] Furthermore, based on the positioning distance between each point in the current circular motion trajectory, the acceleration vector of the acceleration data in the current circular motion trajectory is predicted for the next time period to obtain predicted acceleration data. The predicted target coordinate data is then sampled at the predicted point. The predicted sampling space region corresponding to the predicted target coordinate data is then divided.

[0058] In one implementation, the marginal value d of each axis in the current circular motion trajectory in the sampling space region is obtained by orthogonally covering each node. i Then according to R i =min(r,di ), and obtain the positioning distance R between each point in the current circular motion trajectory i , where r is the standard radius distance of each node. The acceleration data in the current circular motion trajectory is then used with the Lagrangian interpolation function L(i) for the current k-time period to predict the acceleration vector for the k+1 time period. This yields the predicted velocity for each axis, i.e., the x-axis, y-axis, and z-axis, for the k+1 time period. The velocities along the three axes are then vector-added to obtain the acceleration data for each point in the k+1 time period. The x-axis and y-axis are the coordinate axes of the ball in the same plane, and the z-axis is the direction of the wind turbine main shaft.

[0059] Furthermore, based on the signal strength sequence and corresponding density of the predicted points within the predicted sampling space, the predicted motion trend of multiple points within the predicted sampling space is determined. After matching the predicted acceleration data with the predicted points within the predicted sampling space, a predicted circular motion trajectory for the next time period is generated. The circular motion trajectory of the point is then determined based on the predicted circular motion trajectory and the current circular motion trajectory.

[0060] As a feasible implementation method, after obtaining the predicted target coordinate data and the corresponding acceleration data, the predicted target coordinate data corresponding to each sensor chip in the predicted point tracking information is sampled in the next time period to obtain the predicted multi-point position data of the predicted target coordinate data. Then, based on the predicted multi-point position data, the predicted sampling space area corresponding to the point sampling is divided. Through the predicted sampling space area, the predicted multi-point position data is calibrated with respect to the signal strength to obtain the signal strength sequence number of the predicted multi-point. Then, based on the signal strength sequence number and the point density in the predicted sampling space area, the predicted motion trend of the multi-point position data is determined, and based on the matching of the predicted acceleration data obtained above with the predicted point in the predicted sampling space area, a predicted circular motion trajectory based on the next time period is generated.

[0061] S103: Acquire an instantaneous vibration circumferential trajectory of the wind turbine bearing retainer according to the vibration acceleration of the wind turbine bearing retainer.

[0062] Specifically, the vibration sensor in the wind turbine is used to obtain the vibration acceleration in the current time period. Then, the quaternion parameter algorithm is used to perform quaternion differentiation on the motion trend data in the current circular motion trajectory to obtain the operating attitude matrix related to the quaternion differentiation.

[0063] Furthermore, based on the operating posture matrix, the vibration acceleration in the current time period is divided into components of each axis in three-dimensional space to obtain vibration vector coordinate data. The vibration acceleration and vibration vector coordinate data are then transiently fitted to a circular curve to obtain a transient fitting curve.

[0064] Furthermore, according to the three-dimensional space where the wind turbine bearing retainer is located, the transient fitting curve is matched with the corresponding position to determine the instantaneous vibration circular trajectory based on the current time period.

[0065] In one embodiment, the vibration acceleration in the current time period is first obtained based on the vibration sensor in the wind turbine generator set. First, based on the conversion relationship between absolute coordinates and relative coordinates, the motion trend data in the current circular motion trajectory is converted into coordinates using quaternions, that is, the target coordinate data and acceleration data of each point are converted to generate a quaternion differential equation, and then the quaternion differential equation is transformed into a matrix to generate an operating posture matrix related to the quaternion differential.

[0066] In one embodiment, based on the quaternion parameters in the running posture matrix, the vibration acceleration in the current time period is divided into components of the x-axis, y-axis and z-axis to generate a vibration acceleration component matrix of the relevant axis, and then through integration operation, the vibration acceleration component matrix is ​​transiently fitted with the relevant circular curve to obtain a transient fitting curve, and then the transient fitting curve is matched one-to-one with each corresponding point in the three-dimensional space where the current circular motion trajectory is located in the same time period, so that the transient fitting curve and the current circular motion trajectory are in the same circular space area in time and space, and then based on the corresponding vibration acceleration and the transient fitting circular motion trajectory, the instantaneous vibration circular trajectory based on the current time period is determined.

[0067] S104 , filtering the irregular trajectories of the current circular motion trajectory in the circular motion trajectory through the instantaneous vibration circular trajectory to obtain an ideal circular motion trajectory of the wind turbine bearing retainer.

[0068] Specifically, the instantaneous vibration circular trajectory and the current circular motion trajectory are first subjected to linear normalization processing to obtain an instantaneous vibration circular curve and a current circular curve, respectively.

[0069] Furthermore, the instantaneous vibration circular curve and the current circular curve are subjected to difference processing of corresponding coordinate points to obtain the distances of multiple coordinate points, and the distances of multiple coordinate points are subjected to median processing to obtain the vibration difference distance.

[0070] Furthermore, based on the vibration difference distance, the current circular curve is corrected to obtain a corrected circular curve. Based on the acceleration data in the current circular motion trajectory, the corrected circular curve is vector-processed to determine a corrected circular motion trajectory. The corrected circular motion trajectory is then applied to the current circular motion trajectory to filter out irregular trajectories within a preset error range, thereby obtaining the ideal circular motion trajectory for the wind turbine bearing retainer.

[0071] In one embodiment, the instantaneous vibration circular trajectory and the current circular motion trajectory are first linearly normalized to produce an instantaneous vibration circular curve and a current circular curve for easy comparison and calculation. The difference between the corresponding coordinates of the instantaneous vibration circular curve and the current circular curve is then calculated to determine the vibration difference distance at each point. This helps eliminate vibrations in the current circular curve and ensures its uniformity and integrity. The circular motion trajectory is then corrected, and irregular trajectories are filtered within a preset error range to obtain the ideal circular motion trajectory for the wind turbine bearing retainer.

[0072] S105: Generate corresponding ideal circular space regions and corresponding predicted circular space regions for the ideal circular motion trajectory and the predicted circular motion trajectory, and compare the probability centroids of the ideal circular space regions and the predicted circular space regions to obtain predicted overlapping space regions.

[0073] Specifically, based on the ideal circular motion trajectory and the predicted circular motion trajectory, a first spiral cylinder corresponding to the ideal circular space area and a second spiral cylinder corresponding to the predicted circular space area are generated respectively, wherein both the first spiral cylinder and the second spiral cylinder contain multiple point position information.

[0074] Furthermore, the first point position information of the first spiral cylinder is obtained. Then, based on the significance of the probability distribution function, the point distribution plane area corresponding to the first point position information is obtained. The center of mass of the spiral cylinder is located in the point distribution plane area using the probability density function to obtain the first center of mass position information of the first spiral cylinder. The center of mass of the corresponding point distribution plane area of ​​the second spiral cylinder is then located to obtain the second center of mass position information of the second spiral cylinder.

[0075] Furthermore, based on the first center of mass position information and the second center of mass position information, a three-dimensional spatial overlap comparison is performed between the first spiral cylinder and the second spiral cylinder in the same time domain and the same spatial domain to determine the predicted overlapping spatial area that overlaps with the first spiral cylinder and the second spiral cylinder.

[0076] In one embodiment, based on the data representation of the ideal circular motion trajectory and the predicted circular motion trajectory in three-dimensional space, a first spiral cylinder corresponding to the ideal circular space area and a second spiral cylinder corresponding to the predicted circular space area are generated respectively, wherein, based on different time periods, the spatial position and spiral angle of the spiral cylinder are different, and then the first point position information in the first spiral cylinder is significantly identified by the significance of the probability distribution function, and then the point distribution plane area corresponding to the first point position information is obtained. Thereafter, the center of mass of the spiral cylinder is located on the point distribution plane area by the probability density function, the center of mass position data of the first spiral cylinder is identified, and the first center of mass position information of the first spiral cylinder is obtained. Similarly, the center of mass of the relevant point distribution plane area of ​​the second spiral cylinder is continuously located, the center of mass position data of the second spiral cylinder is identified, and finally the second center of mass position information of the second spiral cylinder is obtained.

[0077] As a feasible implementation method, according to the first center of mass position information and the second center of mass position information, the first spiral cylinder and the second spiral cylinder are processed by corresponding movement in the three-dimensional space of the same time domain and the same spatial domain, and then according to each corresponding point position, the first spiral cylinder and the second spiral cylinder are overlapped in spatial volume, and then the mutual overlap predicted overlapping spatial area between the first spiral cylinder and the second spiral cylinder is determined, so as to identify the trajectory portion of the same circular trajectory between the ideal circular motion trajectory and the predicted circular motion trajectory, and eliminate the trajectory portion of the different circular trajectories.

[0078] S106. Based on the predicted overlapping spatial area, determine whether the operation of the wind turbine bearing retainer is abnormal, and obtain operation monitoring information of the wind turbine bearing to complete the operation monitoring of the wind turbine generator set.

[0079] Specifically, a third spiral cylinder in the actual circumferential space region is determined based on the actual circular motion trajectory corresponding to the next time period, wherein the actual circular motion trajectory is the point circular motion trajectory of the next time period based on the ideal circular motion trajectory.

[0080] Furthermore, the center of mass of the third spiral cylinder in the plane area of ​​the relevant point distribution is located to obtain the third center of mass position information of the third spiral cylinder.

[0081] Furthermore, based on the third centroid position information, a three-dimensional spatial overlap comparison is performed between the first spiral cylinder and the third spiral cylinder in the same time domain and the same spatial domain to obtain a true overlap space region.

[0082] Furthermore, by judging the size of the spatial area between the actual overlapping spatial area and the predicted overlapping spatial area, it is determined whether there is any abnormality in the operation of the wind turbine bearing retainer, and the operation monitoring information of the wind turbine bearing is obtained to complete the operation monitoring of the wind turbine group.

[0083] In one embodiment, the actual circular motion trajectory corresponding to the next time period is first obtained, and a third spiral cylinder is constructed for the actual circumferential spatial region, i.e., the spiral cylinder corresponding to the wind turbine bearing retainer's operating trajectory in the next time period. Then, the spatial volume of the third spiral cylinder and the first spiral cylinder in the current time period are compared with each other regarding the center of mass position information to obtain the actual overlapping spatial region between the first and third spiral cylinders, i.e., the actual point-to-point circular motion trajectory of the wind turbine bearing retainer from the current time period to the next time period. The actual overlapping spatial region and the predicted overlapping spatial region are then compared for three-dimensional spatial overlap in the same time and space domains. The higher the overlap, the more consistent the wind turbine bearing retainer's operating conditions are with normal conditions, i.e., the more consistent the wind turbine bearing's operating conditions are with normal operation. Conversely, the lower the overlap, the more the wind turbine bearing retainer's operating conditions deviate from normal operation, the more obvious the abnormality of the wind turbine bearing is, and the more likely it is to cause a potential fault. Finally, the operation monitoring information of the wind turbine bearings is sent to the maintenance personnel through the server in the wind turbine, realizing real-time operation monitoring of the wind turbine. This is conducive to timely detection of potential faults under abnormal operation conditions, allowing maintenance personnel to plan maintenance plans in advance, prevent further increase in wind turbine faults, and solve potential abnormal risks that may cause major faults in advance.

[0084] In addition, the embodiment of the present application further provides an operation monitoring device for a wind turbine bearing retainer. As shown in FIG5 , the operation monitoring device 500 for a wind turbine bearing retainer specifically includes:

[0085] At least one processor 501. And a memory 502 in communication with the at least one processor. The memory 502 stores instructions that can be executed by the at least one processor 501, so that the at least one processor 501 can execute:

[0086] By using multiple sensor chips preset in the wind turbine bearing retainer, the operating status of the wind turbine bearing retainer is tracked at multiple points using multi-cluster head assistance to obtain point tracking information. The point tracking information includes the target coordinate data and corresponding acceleration data of all sensor chips in any time period.

[0087] Performing orthogonal coverage on the point tracking information; and performing motion vector correction and prediction on the point tracking information after orthogonal coverage to obtain a circular motion trajectory of the point; wherein the circular motion trajectory includes: a current circular motion trajectory and a predicted circular motion trajectory;

[0088] Obtaining an instantaneous vibration circumferential trajectory of the wind turbine bearing retainer according to the vibration acceleration of the wind turbine bearing retainer;

[0089] Through the instantaneous vibration circular trajectory, the irregular trajectory of the current circular motion trajectory in the circular motion trajectory is filtered to obtain the ideal circular motion trajectory of the wind turbine bearing retainer;

[0090] Generate corresponding ideal circular space areas and corresponding predicted circular space areas for the ideal circular motion trajectory and the predicted circular motion trajectory respectively; and compare the probability centroids of the ideal circular space areas and the predicted circular space areas to obtain the predicted overlapping space areas;

[0091] Based on the predicted overlapping spatial area, the operation of the wind turbine bearing retainer is judged to determine whether there is any abnormality, and the operation monitoring information of the wind turbine bearing is obtained to complete the operation monitoring of the wind turbine set.

[0092] The beneficial effect of this application is that by monitoring the running trajectory of the wind turbine bearing retainer in the wind turbine main bearing, it can monitor the abnormal operation of the wind turbine bearing retainer in real time based on the error comparison between the predicted overlapping space area and the actual overlapping space area, and then predict whether there are abnormalities in the wind turbine main shaft and the wind turbine group, which is conducive to timely feedback of possible fault problems to maintenance personnel, so that they can arrive at the scene as soon as possible to prevent the further fermentation of potential faults. It is conducive to real-time online supervision and prediction of the operation status of the wind turbine group, reducing the delay time for reporting potential faults of the wind turbine group, facilitating the rapid maintenance of wind turbine groups with abnormalities, reducing maintenance costs, and ensuring the normal power generation efficiency of the wind turbine group.

[0093] The various embodiments in this application are described in a progressive manner. Similar parts between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences from other embodiments. In particular, the device and non-volatile computer storage medium embodiments are generally similar to the method embodiments, so their descriptions are relatively simple. For relevant parts, refer to the descriptions of the method embodiments.

[0094] The foregoing description is of specific embodiments of the present application. In some cases, the actions or steps described in the specification may be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0095] The foregoing is merely an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the embodiments of the present application may have various modifications and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the embodiments of the present application should be included within the scope of the specification of the present application.

Claims

1. A method for monitoring the operation of a wind turbine bearing retainer, characterized in that: The method comprises: Through the multiple sensor chips preset in the wind turbine bearing retainer, the operation status of the wind turbine bearing retainer is tracked with the assistance of multiple cluster heads at multiple points to obtain point tracking information, specifically including: The signal acquisition device in the wind turbine generator set is used to acquire signals from a plurality of sensor chips in the wind turbine bearing holder to determine whether the wind turbine bearing holder is in operation; wherein the plurality of sensor chips are evenly distributed in the wind turbine bearing holder and make the wind turbine bearing holder meet the rotational balance; When the wind turbine bearing retainer is in operation, the signal sending node of the first sensor chip in operation is determined as the main cluster head node, and the signal sending nodes of the second sensor chip and the third sensor chip are determined as adjacent cluster head nodes; wherein the second sensor chip and the third sensor chip are respectively located at left and right adjacent positions of the first sensor chip; According to a preset RSSI algorithm and based on a preset time period, the three-dimensional space distance calculation of the received signal carrier power is performed on the main cluster head node and the adjacent cluster head node respectively, and the main coordinate data and the adjacent coordinate data in the current time period are obtained respectively; wherein the main coordinate data and the adjacent coordinate data are both three-dimensional coordinate data; According to the least squares algorithm, the adjacent coordinate data in the current time period are processed with the minimum value of the average spatial distance, and the median value of the spatial distance after the minimum value processing is calculated to obtain the auxiliary coordinate data; Based on the auxiliary coordinate data and the target coordinate data of the first sensor chip, the point tracking information of all sensor chips is determined; wherein the point tracking information includes: the target coordinate data of all sensor chips in any time period and the corresponding acceleration data; The point tracking information is orthogonally covered, specifically including: Based on the orthogonal coverage mechanism, the target coordinate data corresponding to each sensor chip in the point tracking information is sampled at the current time period to obtain position data of multiple points related to the target coordinate data; According to the position data of the multiple points, a sampling space area corresponding to the sampling of the points is divided; Through the sampling space area, the position data of the multiple points are calibrated with respect to the signal strength, and the signal strength sequence numbers of the multiple points are obtained; According to the signal strength sequence number and the point density in the sampling space area, the movement trend of the position data of the multiple points is determined, and the point movement trend data based on the current time period is determined based on the point with the largest signal strength in the signal strength sequence number as the reference point; Obtain the acceleration data corresponding to each sensor chip in the point tracking signal; and The point motion trend data, one-to-one correspondence between the acceleration data and the position data of multiple points in the sampling space area, to generate a current circular motion trajectory based on the point tracking information; and the point tracking information after orthogonal coverage is corrected and predicted for the motion vector of the relevant point to obtain the circular motion trajectory of the point, specifically including: Get the point tracking information after orthogonal coverage in the current time period; According to the Lagrange interpolation function and based on the acceleration data in the current circular motion trajectory, the coordinate position vector of the target coordinate data in the current circular motion trajectory is predicted for the next time period to obtain the predicted target coordinate data; According to the positioning distance between each point in the current circular motion trajectory, the acceleration vector of the next time period is predicted for the acceleration data in the current circular motion trajectory to obtain predicted acceleration data; Sampling the predicted points of the predicted target coordinate data; and dividing the predicted sampling space area corresponding to the predicted target coordinate data; According to the signal strength sequence number of the predicted point in the predicted sampling space area and the corresponding predicted point density, the predicted motion trend of multiple points in the predicted sampling space area is determined; and according to the predicted point in the predicted sampling space area and the corresponding predicted acceleration data, a predicted circular motion trajectory based on the next time period is generated; Based on the predicted circular motion trajectory and the current circular motion trajectory, the circular motion trajectory of the point is obtained; wherein the circular motion trajectory includes: the current circular motion trajectory and the predicted circular motion trajectory; According to the vibration acceleration of the wind turbine bearing retainer, obtaining the instantaneous vibration circumferential trajectory of the wind turbine bearing retainer specifically includes: The vibration acceleration in the current time period is obtained through the vibration sensor in the wind turbine; Through the quaternion parameter algorithm, the motion trend data in the current circular motion trajectory is divided into quaternion differentials to obtain the running posture matrix related to the quaternion differentials; According to the running posture matrix, the vibration acceleration in the current time period is divided into components of each axis in the three-dimensional space to obtain vibration vector coordinate data; Performing transient fitting of a circular curve on the vibration acceleration and the vibration vector coordinate data to obtain a transient fitting curve; Based on the three-dimensional space where the wind turbine bearing retainer is located, matching the transient fitting curve with the corresponding position to determine the instantaneous vibration circular trajectory based on the current time period; By using the instantaneous vibration circular trajectory, filtering the irregular trajectory of the current circular motion trajectory in the circular motion trajectory to obtain the ideal circular motion trajectory of the wind turbine bearing retainer, specifically including: The instantaneous vibration circular trajectory and the current circular motion trajectory are subjected to linear normalization processing to obtain an instantaneous vibration circular curve and a current circular curve respectively; wherein the instantaneous vibration circular curve and the current circular curve are both spiral circular curves; Performing difference processing on corresponding coordinate points of the instantaneous vibration circular curve and the current circular curve to obtain distances of multiple coordinate points; and performing median processing on the distances of the multiple coordinate points to obtain a vibration difference distance; According to the vibration difference distance, the current circular curve is subjected to curve correction to obtain a corrected circular curve; and according to the acceleration data in the current circular motion trajectory, the corrected circular curve is subjected to vector processing to determine a corrected circular motion trajectory; By using the modified circular motion trajectory, the current circular motion trajectory is filtered and screened for irregular trajectories within a preset error range, so as to obtain an ideal circular motion trajectory of the wind turbine bearing retainer; Generating a corresponding ideal circular space region and a corresponding predicted circular space region for the ideal circular motion trajectory and the predicted circular motion trajectory respectively; and performing a positioning comparison of the probability centroid of the ideal circular space region and the predicted circular space region to obtain a predicted overlapping space region, specifically comprising: According to the ideal circular motion trajectory and the predicted circular motion trajectory, a first spiral cylinder corresponding to the ideal circular space area and a second spiral cylinder corresponding to the predicted circular space area are generated respectively; wherein the first spiral cylinder and the second spiral cylinder both contain a plurality of point position information; Acquire the position information of a first point in the first spiral cylinder; According to the significance of the probability distribution function, a point distribution plane area corresponding to the first point position information is obtained; and the center of mass of the spiral cylinder is located in the point distribution plane area through the probability density function to obtain the first center of mass position information of the first spiral cylinder; Positioning the center of mass of the second spiral cylinder in the relevant point distribution plane area to obtain second center of mass position information of the second spiral cylinder; According to the first mass center position information and the second mass center position information, a three-dimensional spatial coincidence comparison is performed between the first spiral cylinder and the second spiral cylinder in the same time domain and the same space domain to determine a predicted coincidence spatial region that coincides with the first spiral cylinder and the second spiral cylinder; According to the predicted overlap space area, whether there is any abnormality in the operation of the wind turbine bearing retainer is judged, and the operation monitoring information of the wind turbine bearing is obtained, specifically including: According to the actual circular motion trajectory corresponding to the next time period, a third spiral cylinder related to the actual circular space area is determined; wherein the actual circular motion trajectory is a point circular motion trajectory of the next time period based on the ideal circular motion trajectory; Positioning the centroid of the third spiral cylinder in the relevant point distribution plane area to obtain third centroid position information of the third spiral cylinder; According to the third centroid position information, a three-dimensional spatial overlap comparison is performed between the first spiral cylinder and the third spiral cylinder in the same time domain and the same space domain. Get the real overlapping space area; By judging the spatial area size information of the real overlapping spatial area and the predicted overlapping spatial area, it is judged whether there is any abnormality in the operation of the wind turbine bearing retainer, and the operation monitoring information of the wind turbine bearing is obtained to complete the operation monitoring of the wind turbine set.

2. The operation monitoring method of a wind turbine bearing retainer according to claim 1, characterized in that: Based on the auxiliary coordinate data and the target coordinate data of the first sensor chip, the point tracking information of all sensor chips is determined, specifically including: Based on the current time period, the auxiliary coordinate data in the adjacent cluster head node and the main coordinate data in the main cluster head node are subjected to weight fusion of the relevant coordinate data to obtain the target coordinate data of the first sensor chip; According to a preset WSN grid structure, the signal sending node of the first sensor chip is determined as an adjacent cluster head node, and the target coordinate data is determined as adjacent coordinate data; The target coordinate data of the second sensor chip is obtained by performing minimum value processing of the average spatial distance on the coordinate data of the fourth sensor chip and the target coordinate data of the first sensor chip, and then fusing the coordinate data based on the weights; wherein the fourth sensor chip and the first sensor chip are both placed in adjacent positions on the left and right of the second sensor chip; Similarly, weighted fusion of the relevant coordinate data of all sensor chips is performed to determine the target coordinate data of all sensor chips respectively; and the acceleration data corresponding to the multiple sensor chips in the current time period is obtained; Based on the target coordinate data of all sensor chips and the corresponding acceleration data, the point tracking information of all sensor chips is determined.

3. An operation monitoring device for a wind turbine bearing retainer, characterized in that: The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, so that the at least one processor can execute the operation monitoring method of a wind turbine bearing retainer according to any one of claims 1-2.