Wind power bolt real-time stress monitoring method and system

Abstract

The invention relates to stress monitoring, in particular to a real-time stress monitoring method and system for a wind power bolt, and the method comprises the steps: 1, collecting the rotating speed of a main shaft of a motor, the voltage of the output end of the motor and air pressure data of a plurality of positions on the surface of a blade according to a preset frequency; 2, based on the collected motor main shaft rotating speed, the motor output end voltage and the air pressure, the dynamic total load of the motor is obtained; 3, distributing the dynamic total load to the bolts at different positions through a total load distribution model, namely mapping the dynamic total load into the stress of the bolts at different positions; 4, monitoring the stress of the bolts at different positions, judging whether the bolts have faults or not, if yes, entering the step 5, and otherwise, returning to the step 1; and 5, carrying out bolt replacement processing, analyzing and evaluating the influence of the processed bolt on the stress of other bolts, and judging the stress safety of other bolts.

Description

Technical Field

[0001] The present invention relates to the technical field of stress monitoring, and particularly relates to a real-time stress monitoring method and system for wind power bolts. Background Art

[0002] During the operation of a wind turbine, it needs to withstand wind loads, blade centrifugal forces, etc. The randomness of wind loads causes the bolts in the wind turbine to bear periodic tensile and compressive forces (such as periodic vibrations when the blades rotate), that is, the bolts need to bear dynamic loads. The torque of the gearbox and the swing of the blades will generate lateral forces on the bolts, and the friction of the bolt connection surface will cause the bolts to bear a certain shear force. If the bolt connection surface is uneven or there is an installation deviation, it will cause the bolts to be eccentrically stressed. In addition, the tower flange will undergo local bending under strong winds, forcing the bolts to bear bending moments. That is to say, the stress states of bolts at different positions on the wind turbine are the result of the coupling of multiple physical fields. Real-time monitoring of bolt stress is an important means to ensure the safe operation of wind turbine equipment. Existing technical means are mainly divided into direct measurement methods (such as the resistance strain method based on strain gauges and the ultrasonic propagation time method, etc.) and indirect inference methods (such as the method of back-calculating based on pre-tightening force and torque). These two types of methods have defects such as low environmental adaptability, high cost, and poor monitoring accuracy.

[0003] Therefore, through AI technology, combining stress data and the operating state of the wind turbine, monitoring bolt stress and predicting bolt failures through machine learning algorithms has become a development trend. Summary of the Invention

[0004] Based on the motor operating data such as the rotational speed of the motor and the output voltage, and combining with the air pressure received by the blade, the present invention indirectly obtains the dynamic total load; and distributes the dynamic total load to different bolts to obtain the stress data of different bolts, realizing the function of monitoring the stress of bolts.

[0005] The technical solution proposed by the present invention is: a real-time stress monitoring method for wind power bolts, the method comprising: Step 1: Collect data on the rotational speed of the motor spindle, the output voltage of the motor, and the air pressure at multiple positions on the blade surface at a preset frequency; Step 2: Obtain the dynamic total load of the motor based on the collected rotational speed of the motor spindle, the output voltage of the motor, and the air pressure; Step 3: Distribute the dynamic total load to bolts at different positions through a total load distribution model, that is, map the dynamic total load to the stress of bolts at different positions; Step 4: Monitor the stress of bolts at different positions, determine whether the bolts are faulty. If so, enter Step 5; otherwise, return to Step 1; Step 5: Conduct bolt replacement treatment, analyze and evaluate the impact of the treated bolts on the stress of other bolts, and determine the stress safety of other bolts.

[0006] Preferably, obtaining the dynamic total load of the motor based on the collected motor spindle speed, motor output voltage, and air pressure includes the following steps: Convert the motor spindle speed to angular velocity , and obtain the current corresponding to the motor output voltage , and calculate the electromagnetic torque , where represents the output voltage of the motor; Obtain the aerodynamic torque of the motor blades based on the pressure data , where respectively represent the pressure at the blade surface position , the radius of rotation at position , and the wind energy utilization coefficient; represents the total number of positions; Calculate the mechanical loss torque compensation , where represents the friction coefficient, represents the cogging torque; Then the dynamic total load is ; Allocating the dynamic total load to bolts at different positions through the total load distribution model, that is, mapping the dynamic total load to the stress of bolts at different positions, includes: Establish a load transfer path to transfer the dynamic total load to each bolt; specifically including: Apply a unit force or a unit moment to the blade in sequence, and record the response force or moment of each bolt; Calculate the unit axial force distribution coefficient of the bolt based on the response force of each bolt and the applied unit force or moment; Use the unit distribution coefficient to construct a dynamic total load distribution matrix ; where represents the row vector of the axial force distribution coefficient; represents the row vector of the bending moment distribution coefficient; Construct a total load distribution model ; where respectively represent the axial tensile force, shear force, and torque allocated to the th bolt; respectively represent the force and torque of the dynamic total load in the -axis direction in the global coordinate system.

[0007] Preferably, monitoring the stress of bolts at different positions to determine whether the bolts are faulty includes: Using the pre-trained regression model LSTM to monitor the bolt stress and determine whether the bolts are faulty; The input of the LSTM is the axial tensile force, shear force and torque of different bolts after normalization processing, and the output is the combined stress of different bolts; Set the bolt fault threshold to 70% of the maximum yield strength of the bolt; If the combined stress is greater than the bolt fault threshold, it is determined that the bolt is faulty and needs to be replaced; otherwise, it does not need to be replaced.

[0008] Preferably, distributing the dynamic total load to the bolts at different positions through the total load distribution model, that is, mapping the dynamic total load to the stress of the bolts at different positions, further includes: Considering the force transfer between bolts, introducing a coupling matrix Optimizing the dynamic total load distribution matrix to obtain the optimized dynamic total load matrix ; where represents the dynamic total load distribution matrix of bolt , and the element in represents the force transfer coefficient of bolt to bolt ; , where represents the local stiffness of bolt , represents the coupling stiffness of bolts and ;

[0009] Preferably, distributing the dynamic total load to the bolts at different positions through the total load distribution model, that is, mapping the dynamic total load to the stress of the bolts at different positions, further includes: Dividing the dynamic total load into aerodynamic load and mechanical load; the mechanical load includes one or more of gearbox torque, inertial load of rotating components, gravity load and friction loss; Mechanical load ; Setting the aerodynamic load distribution matrix, calibrated through wind tunnel experiments, specifically including: Applying a stepped wind speed in the wind tunnel and measuring the strain of each bolt and the aerodynamic load vector in the global coordinate system; Aerodynamic load distribution matrix ; Setting the mechanical load distribution matrix, calibrated through bench tests, specifically including: Disconnect the blade and apply a stepped torque through the motor; Measure the strain of each bolt and the mechanical load vector in the global coordinate system ; Mechanical load distribution matrix ; Using the aerodynamic load distribution matrix and the aerodynamic load vector in the global coordinate system, calculate the axial tensile force, shear force and torque of the bolt, and then obtain the combined stress under the aerodynamic load; if the combined stress under the aerodynamic load exceeds the preset aerodynamic load combined stress threshold, it is determined that the abnormal bolt stress is caused by the aerodynamic load; Using the mechanical load distribution matrix and the mechanical load vector in the global coordinate system, calculate the axial tensile force, shear force and torque of the bolt, and then obtain the combined stress under the mechanical load; if the combined stress under the mechanical load exceeds the preset mechanical load combined stress threshold, it is determined that the abnormal bolt stress is caused by the mechanical load.

[0010] Preferably, the monitoring of the stress of bolts at different positions to determine whether the bolts are faulty further includes: constructing a state space model, capturing the interaction between the mechanical load and the aerodynamic load through the state space model, and identifying the stress distribution of the bolts, specifically including the following steps: Constructing a state space model includes: Obtain the deformation of the bolt connection part , that is, the rigidity of the bolt; Obtain the bolt vibration velocity , and the strain of the bolt ; Construct a state vector ; State equation ; where respectively represent the state transition matrix, input matrix and observation matrix; Input vector , where respectively represent the wind speed, the pressure at the th position on the blade surface and the gearbox torque; By using the system identification method N4SID, determine the parameters of the matrix ; through the off-diagonal elements in reflect the dynamic coupling between state vectors; through the elements in reflect the influence degree of the elements in the input vector on the elements in the state vector; Identifying the stress distribution of the bolt includes: Associate the state vector with the bolt stress through the observation matrix; if only the bolt stress is concerned, then , denotes the identity matrix; Using the model after system identification method, input the real-time load data to predict the bolt stress: , where respectively represent the state vector at the starting moment and the input vector at the moment; Obtain the measured stress of the bolt, calculate the difference between the predicted bolt stress and the measured stress, that is, the stress residual ; Calculate the correlation coefficients between the stress residuals of each bolt and the corresponding mechanical load distribution matrix and aerodynamic load distribution matrix, specifically: Aerodynamic load correlation coefficient , mechanical load correlation coefficient ; If then it is judged that the bolt failure comes from the aerodynamic load; If then it is judged that the bolt failure comes from the mechanical load; Obtain the real-time wind speed data and the air pressure at multiple positions on the blade surface, judge whether the air pressure matches, so as to judge whether the variables related to the aerodynamic load in the input vector are normal; measure the vibration data of the gearbox, extract the frequency spectrum data when the gears are meshing, and detect whether the frequency assignment increases suddenly when the gears are meshing, so as to judge whether the variables related to the mechanical load in the input vector are normal; If the variables related to the mechanical load and the aerodynamic load are both normal, but the residuals are continuously abnormal, then it is judged that the mechanical load distribution matrix and the aerodynamic load distribution matrix fail, and the mechanical load distribution matrix and the aerodynamic load distribution matrix need to be recalibrated.

[0011] Preferably, the bolt replacement treatment is carried out, and the influence of the treated bolt on the stress of other bolts is analyzed and evaluated, including: Obtain the bolt data after the fault recovery, including the new pre-tightening force and the new stiffness of the bolt; The new stiffness of the bolt ; where, respectively represent the new pre-tightening force and the bolt elongation; Update the bolt For the bolt coupling coefficient, obtain the new coupling coefficient matrix ; where, represents the bolt before replacement axial stiffness; Obtain the historical pre-tightening force of the faulty bolt from the database , update the dynamic total load distribution matrix, and obtain the updated dynamic total load distribution matrix ; Using and Calculate the new axial tensile force, shear force and torque, and then obtain the new stress matrix ; Calculate the tensile stress using the stress matrix , shear stress and bending stress ; where represents the distance from the neutral axis of the bolt cross-section to the outermost edge. If the cross-section of the bolt is circular, then , represents the diameter of the cross-section; The combined stress is ; The change rate of the combined stress , if , then it is determined to update the bolt affects the stress safety of other bolts.

[0012] Preferably, the bolt replacement process is carried out, and the influence of the treated bolt on the stress of other bolts is analyzed and evaluated, and further includes: Consider the influence of the installation of the new bolt on the local contact pressure distribution, the influence of the pre-tightening force of the new bolt on the structural deformation of its contact part, and the change of the friction coefficient under dynamic load. Specifically: Measure the axial stiffness of the bolt using a hydraulic tensioner; Measure the contact pressure distribution of the bolt connection surface using an ultrasonic detector and calculate the friction coefficient ; Monitor the pre-tightening force attenuation ratio of adjacent bolts through distributed strain gauges ; Correct the coupling coefficient to obtain the corrected new coupling coefficient , where represents the bolt belongs to the adjacent bolt set ; respectively represent the friction coefficients of the bolts before and after replacement; Apply a stepped load to the bolt and record the measured strain response; Construct a nonlinear regression model , where represents the measured strain response, that is, the measured stress matrix , represents the stepped load vector; represents the coefficient matrix; , ; Output the prediction through a non - linear regression model and combine with to calculate and obtain the corresponding tensile stress, shear stress and bending stress, and then obtain the corresponding synthetic stress change rate to judge the impact of the installation of the new bolt on the stress safety of other bolts.

[0013] The present invention also provides a computer - readable storage medium storing a computer program, and the computer program is executed by a processor to implement the real - time stress monitoring method for wind power bolts.

[0014] Advantages of the present invention: 1. In the present invention, by collecting the rotational speed of the fan, the air pressure at different positions on the blade surface, and the voltage data at the output end, establishing the association between these data and bolts at different positions, calculating the force and moment transmitted to the bolts through different loads by the load distribution model, and predicting the stress of the bolts according to the force and moment. By analyzing the stress residuals between the measured stress and the predicted stress, it is judged whether the bolt is faulty.

[0015] 2. In the present invention, based on the rotational speed of the fan, the air pressure at different positions on the blade surface, and the voltage data at the output end, the dynamic total load is obtained, and the dynamic total load is divided into mechanical load (which can be applied to the monitoring of bolts in the gearbox) and aerodynamic load (which can be used for the stress monitoring of bolts at the blade root). And by calculating the correlation coefficients between the stress residuals of each bolt and the mechanical load and aerodynamic load distribution matrices, it is judged whether the source of the bolt fault is the aerodynamic load or the mechanical load. Description of the Drawings

[0016] Figure 1 is a flowchart of the real - time stress monitoring method for wind power bolts of the present invention. Detailed Embodiments

[0017] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments in the following description are only examples, and those skilled in the art can think of other obvious deformations. The basic principles defined in the following description can be applied to other implementation schemes, deformation schemes, improvement schemes, equivalent schemes, and other technical schemes without departing from the spirit and scope of the present invention.

[0018] It can be understood that the term "a" should be understood as "at least one" or "one or more". That is, in one embodiment, the number of an element can be one, while in other embodiments, the number of the element can be multiple. The term "a" cannot be understood as a limitation on the number.

[0019] Embodiment 1: Refer to Figure 1, the technical solution provided by the present invention is: a real-time stress monitoring method for wind power bolts, including the following steps: Step 1: Collect the motor spindle speed, the voltage at the motor output end, and the air pressure data at multiple positions on the blade surface at a preset frequency; Step 2: Obtain the dynamic total load of the motor based on the collected motor spindle speed, the voltage at the motor output end, and the air pressure; including: Step 2.1: Convert the motor spindle speed to angular velocity , obtain the current corresponding to the voltage at the motor output end , calculate the electromagnetic torque , where represents the output voltage of the motor; Step 2.2: Obtain the aerodynamic torque of the motor blade based on the pressure data , where respectively represent the pressure at the position on the blade surface, the radius of rotation at the position , and the wind energy utilization coefficient; represents the total number of positions; Step 2.3: Calculate the mechanical loss torque compensation , where represents the friction coefficient, represents the cogging torque; Then the dynamic total load is ; Step 3: Allocate the dynamic total load to the bolts at different positions through the total load distribution model, that is, map the dynamic total load to the stresses of the bolts at different positions, including the following steps: Establish a load transfer path to transfer the dynamic total load to each bolt, specifically including: Apply a unit force or a unit moment to the blade in sequence, and record the response force or moment of each bolt; Calculate the unit axial force distribution coefficient of the bolt based on the response force of each bolt and the applied unit force or moment; Construct a dynamic total load distribution matrix using the unit distribution coefficient ; where represents the row vector of the axial force distribution coefficient; represents the row vector of the bending moment distribution coefficient; Construct the total load distribution model ; where respectively represent the axial tensile force, shear force, and torque allocated to the th bolt; respectively represent the force and torque of the dynamic total load in the axial direction in the global coordinate system.

[0020] Step 4: Monitor the stress of bolts at different positions to determine whether the bolts are faulty. If so, go to Step 5; otherwise, return to Step 1. Specifically: Use the pre-trained regression model LSTM to monitor the bolt stress and determine whether the bolts are faulty; The input of the LSTM is the axial tensile force, shear force and torque of different bolts after normalization processing, and the output is the combined stress of different bolts; Set the bolt fault threshold to 70% of the maximum yield strength of the bolt; If the combined stress is greater than the bolt fault threshold, it is determined that the bolt is faulty and needs to be replaced; otherwise, it does not need to be replaced.

[0021] For example, through measurement, KN, the torque in the axial direction is KNm; ; ; Then, the tensile stress MPa (taking M36 bolt as an example, the effective cross-sectional area , the moment of inertia of the bolt cross-section , the yield strength MPa).

[0022] The shear stress MPa; The bending stress MPa; The combined stress MPa; Set the safety threshold to 900×0.7 = 630 MPa. Since 28 MPa is less than 630 MPa, it is determined that the bolt stress is normal, the bolt is safe and does not need to be replaced.

[0023] Step 5: Perform bolt replacement treatment, analyze and evaluate the influence of the treated bolts on the stress of other bolts, and judge the stress safety of other bolts, including the following steps: Obtain the bolt data after fault recovery, including the new pre-tightening force and new stiffness of the bolts; The new stiffness of the bolt ; where respectively represent the new pre-tightening force and bolt elongation; Update the bolt For the bolt The coupling coefficient to obtain a new coupling coefficient matrix ; where represents the bolt before replacement axial stiffness; Obtain the historical pre-tightening force of the faulty bolt from the database , update the dynamic total load distribution matrix to obtain the updated dynamic total load distribution matrix ; Use and to calculate the new axial tensile force, shear force and torque, and then obtain the new stress matrix ; Calculate and obtain the tensile stress using the stress matrix , shear stress and bending stress ; where represents the distance from the neutral axis of the bolt cross-section to the outermost edge. If the cross-section of the bolt is circular, then , represents the diameter of the cross-section; The combined stress is ; The change rate of the combined stress , if , then determine that the updated bolt affects the stress safety of other bolts.

[0024] Example 2: When distributing the total load to bolts at different positions, the mutual influence between bolts needs to be considered. And when replacing new bolts, the influence of the replacement of new bolts on other bolts also needs to be considered. For this reason, the following scheme is further proposed on the basis of Example 1: Consider the force transfer between bolts and introduce a coupling matrix to optimize the dynamic total load distribution matrix and obtain the optimized dynamic total load matrix ; where represents the dynamic total load distribution matrix of bolt , and the elements in represent the force transfer coefficient of bolt to bolt ; , where represents the local stiffness of bolt , represents the coupling stiffness of bolt and ;

[0025] ​Consider the influence of the installation of new bolts on the local contact pressure distribution, the influence of the pre-tightening force of new bolts on the structural deformation of their contact parts, and the change of the friction coefficient under dynamic loads, specifically: Measure the axial stiffness of the bolts through a hydraulic tensioner ; ; Measure the contact pressure distribution of the bolt connection surface through an ultrasonic detector and calculate the friction coefficient ; Monitor the pre-tightening force attenuation ratio of adjacent bolts through distributed strain gauges ; Modify the coupling coefficient to obtain a new modified coupling coefficient, where , represents the bolt belonging to the set of adjacent bolts ; respectively represent the friction coefficients of the bolts before and after replacement ; Apply a stepped load to the bolt and record the measured strain response; Construct a nonlinear regression model , where represents the measured strain response, that is, the measured stress matrix , represents the stepped load vector; represents the coefficient matrix; , ; Output the predicted through the nonlinear regression model, and combine to calculate and obtain the corresponding tensile stress, shear stress and bending stress, and then obtain the corresponding synthetic stress change rate to judge the influence of the installation of new bolts on the stress safety of other bolts.

[0026] Example 3: When distributing the dynamic total load to bolts at different positions through the total load distribution model, the technical solution given in Example 1 is to distribute the load as a whole without considering the separate distribution of the mechanical load and the pneumatic load included in the total load. Therefore, on the basis of Example 1, we further improve the total load distribution method and propose the following technical solution: Divide the dynamic total load into pneumatic load and mechanical load; the mechanical load includes one or more of gearbox torque, inertial load of rotating parts, gravity load and friction loss; Mechanical load ; Set up a pneumatic load distribution matrix and calibrate it through a wind tunnel experiment, specifically including: Apply a stepped wind speed in the wind tunnel and measure the strain of each bolt and the aerodynamic load vector in the global coordinate system ; Aerodynamic load distribution matrix ; Set the mechanical load distribution matrix and calibrate it through bench tests, specifically including: Disconnect the blade and apply a stepped torque through the motor; Measure the strain of each bolt and the mechanical load vector in the global coordinate system ; Mechanical load distribution matrix ; Use the aerodynamic load distribution matrix and the aerodynamic load vector in the global coordinate system to calculate the axial tensile force, shear force and torque of the bolt, and then obtain the combined stress under the aerodynamic load; if the combined stress under the aerodynamic load exceeds the preset aerodynamic load combined stress threshold, it is determined that the abnormal bolt stress is caused by the aerodynamic load; Use the mechanical load distribution matrix and the mechanical load vector in the global coordinate system to calculate the axial tensile force, shear force and torque of the bolt, and then obtain the combined stress under the mechanical load; if the combined stress under the mechanical load exceeds the preset mechanical load combined stress threshold, it is determined that the abnormal bolt stress is caused by the mechanical load.

[0027] For example, in this embodiment, through measurement, it is obtained that: the wind speed is 12 m / s, the motor spindle speed is 15 rpm, the output terminal voltage is 690 V, the current is 200 A, ; the blade length is 40 m.

[0028] Then the aerodynamic load is: KNm; wherein, respectively represent air density, wind speed, wind energy utilization coefficient, blade swept area and angular velocity; Electromagnetic torque KNm; Mechanical loss KNm; Mechanical load KNm; The bolt load is distributed as mechanical load distribution and aerodynamic load distribution: and ; By Calculate the strain of the corresponding bolt under the aerodynamic load. If the corresponding strain exceeds the preset aerodynamic strain threshold, it is determined that the bolt failure is caused by the aerodynamic load; By Calculate the strain of the corresponding bolt under mechanical load. If the corresponding strain exceeds the preset mechanical strain threshold, it is determined that the bolt failure is caused by mechanical load. Since the pneumatic load is a periodic alternating force generated by the wind acting on the blade, it is likely to cause bolt fatigue failure. Mechanical loads include gearbox torque, inertial force, and friction loss, etc., which are likely to cause bolt overload or loosening. Therefore, by separately distributing the loads and monitoring the corresponding stresses, it helps to determine the type of bolt failure that is about to occur.

[0029] Example 4: In Example 3, we proposed a technical solution for separately distributing the total load. Based on this technical solution, when judging bolt failure and identifying bolt stress distribution, the interaction between mechanical load and pneumatic load needs to be considered. Therefore, we propose the following solution: Construct a state space model to capture the interaction between mechanical load and pneumatic load through the state space model and identify the stress distribution of the bolt, which specifically includes the following steps: Construct a state space model, including: Obtain the deformation of the bolt connection part , that is, the rigidity of the bolt; Obtain the bolt vibration velocity , and the strain of the bolt ; Construct a state vector ; State equation ; where respectively represent the state transition matrix, input matrix, and observation matrix; Input vector , where respectively represent the wind speed, the pressure at the th position on the blade surface, and the gearbox torque; By using the system identification method N4SID, determine the parameters of the matrix ; through the off-diagonal elements in reflect the dynamic coupling between state vectors; through the elements in reflect the influence degree of the elements in the input vector on the elements in the state vector.

[0030] For example, in the real-time stress monitoring of the gearbox bolt group of an onshore wind turbine, the state vector includes the vibration displacement, velocity of the gearbox, and the stress of the bolt ; The input vector includes the wind speed and gearbox torque, and outputs the stress data of each bolt in the bolt group; Collect 10 groups of working condition data and identify using the N4SID method to obtain: , ; , through the matrix it is shown that the wind speed (aerodynamic) has a greater impact on State 1 (vibration displacement), and the gearbox torque (mechanical) has a greater impact on State 2 (speed). The coupling term indicates that the speed change affects the displacement, and then through it affects the stress of the bolt. Thus, it reflects the influence of the aerodynamic load on the mechanical load state and the influence on the stress after coupling.

[0031] Identify the stress distribution of the bolt, including: Through the observation matrix, correlate the state vector with the bolt stress; if only focusing on the bolt stress, then ; Use the model after the system identification method, input the real-time load data to predict the bolt stress: , where respectively represent the state vector at the starting moment and the input vector at the moment; Obtain the measured stress of the bolt, calculate the difference between the predicted bolt stress and the measured stress, that is, the stress residual ; Calculate the correlation coefficients of the stress residuals of each bolt with the corresponding mechanical load distribution matrix and aerodynamic load distribution matrix, specifically: Aerodynamic load correlation coefficient , mechanical load correlation coefficient ; If then it is judged that the failure of the bolt comes from the aerodynamic load; If then it is judged that the failure of the bolt comes from the mechanical load; Obtain the real-time wind speed data and the air pressure at multiple positions on the blade surface, judge whether the air pressure matches to judge whether the variables related to the aerodynamic load in the input vector are normal; measure the vibration data of the gearbox, extract the frequency spectrum data when the gears are meshing, and detect whether the frequency assignment increases suddenly when the gears are meshing to judge whether the variables related to the mechanical load in the input vector are normal; If the variables related to both the mechanical load and the aerodynamic load are normal, but the residuals are continuously abnormal, then it is judged that the mechanical load distribution matrix and the aerodynamic load distribution matrix fail, and the mechanical load distribution matrix and the aerodynamic load distribution matrix need to be recalibrated.

[0032] The present invention also provides a real-time stress monitoring system for wind power bolts, comprising: a measurement module, a monitoring module and a communication module. Among them, the measurement module is used to collect the rotational speed of the motor main shaft, the air pressure at multiple positions on the blade surface, as well as the motor output voltage and current. The monitoring module is used to analyze and process the data input by the measurement module, monitor the stress of different screws, judge whether different bolts are faulty, and output the judgment result. The communication module is used to connect to the host computer and transmit the judgment result to the host computer; the system is used to execute the above-mentioned real-time stress monitoring method for wind power bolts. Specifically, the measurement module includes a rotational speed sensor, a plurality of pressure sensors and a voltage transformer, and the monitoring module includes a processor and a memory connected to the processor.

[0033] The present invention also provides a computer-readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the above-mentioned real-time stress monitoring method for wind power bolts.

[0034] Embodiments disclosed by the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. Embodiments disclosed by the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for performing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from a network through a communication part, and / or installed from a removable medium. When the computer program is executed by a central processing unit (CPU), the above-mentioned functions defined in the methods of the present invention are performed. It should be noted that the computer-readable medium mentioned above in the present invention can be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. A computer-readable storage medium can, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of a computer-readable storage medium can include, but are not limited to: an electrical connection having one or more wire segments, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, apparatus, or device. And in the present invention, a computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program codes. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, and this computer-readable medium can send, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device. The program codes contained on the computer-readable medium can be transmitted by any suitable medium, including but not limited to: wireless segments, wire segments, optical cables, RF, etc., or any suitable combination of the above.

[0035] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a part of code, which contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0036] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are only examples and do not limit the present invention. The objectives of the present invention have been fully and effectively achieved. The functions and structural principles of the present invention have been demonstrated and explained in the embodiments. Without departing from the said principles, any changes or modifications can be made to the embodiments of the present invention.

Claims

1. A wind power bolt real-time stress monitoring method, characterized in that: The method comprises: Step 1: collecting the motor spindle speed, the motor output terminal voltage and the air pressure data at multiple positions on the blade surface according to a preset frequency; Step 2: Obtain the total dynamic load of the motor based on the collected motor spindle speed, motor output terminal voltage and air pressure; Step 3: Distribute the dynamic total load to bolts at different positions through the total load distribution model, that is, map the dynamic total load to the stress of bolts at different positions; Step 4: Monitor the stress of bolts at different positions to determine whether the bolts are faulty. If so, proceed to step 5; otherwise, return to step 1. Step 5: Replace the bolts, analyze and evaluate the effect of the replaced bolts on the stress of other bolts, and determine the stress safety of other bolts.

2. A wind power bolt real-time stress monitoring method according to claim 1, characterized in that: The method of obtaining the total dynamic load of the motor based on the collected motor spindle speed, motor output terminal voltage and air pressure comprises the following steps: Convert motor shaft speed to angular velocity , get the current corresponding to the motor output voltage , calculate the electromagnetic torque ,in, Indicates the output voltage of the motor; Obtaining aerodynamic torque of motor blades based on pressure data ,in, Represents the blade surface position Pressure, position The rotation radius, wind energy utilization coefficient; Indicates the total number of positions; Calculation of mechanical loss torque compensation ,in, represents the friction coefficient, represents the cogging torque; The total dynamic load is ; The method of distributing the dynamic total load to bolts at different positions by using the total load distribution model, that is, mapping the dynamic total load to the stress of bolts at different positions, includes: Establish a load transfer path to transfer the total dynamic load to each bolt; specifically include: Apply unit force to the blades one by one or unit moment , record the response force of each bolt or torque ; Calculate the unit axial force distribution coefficient of the bolt based on the response force of each bolt and the applied unit force or moment; Constructing a dynamic total load distribution matrix using unit distribution coefficients ;in, represents the row vector of axial force distribution coefficient; represents the row vector of the moment distribution coefficient; Constructing the total load distribution model ;in, Respectively indicate the allocation to Axial tension, shear and torque of each bolt; They represent the total dynamic load in the global coordinate system. Forces and torques in the axial direction.

3. A wind power bolt real-time stress monitoring method according to claim 2, characterized in that: The monitoring of the stress of bolts at different positions to determine whether the bolts are faulty includes: Use the pre-trained regression model LSTM to monitor bolt stress and determine whether the bolt is faulty; The input of the LSTM is the normalized axial tensile force, shear force and torque of different bolts, and the output is the composite stress of different bolts; The bolt failure threshold is set to 70% of the maximum yield strength of the bolt; If the composite stress is greater than the bolt failure threshold, the bolt is judged to be faulty and needs to be replaced; otherwise, it does not need to be replaced.

4. A wind power bolt real-time stress monitoring method according to claim 3, characterized in that: The method of distributing the dynamic total load to bolts at different positions by using the total load distribution model, that is, mapping the dynamic total load to the stress of bolts at different positions, further includes: Considering the force transmission between bolts, the coupling matrix is ​​introduced Optimize the dynamic total load distribution matrix to obtain the optimized dynamic total load matrix ;in, Indicates bolt The dynamic total load distribution matrix, The elements in the bolt Bolt The force transmission coefficient; ,in, Indicates bolt The local stiffness of Indicates bolt and The coupling stiffness.

5. A wind power bolt real-time stress monitoring method according to claim 4, characterized in that: The method of distributing the dynamic total load to bolts at different positions by using the total load distribution model, that is, mapping the dynamic total load to the stress of bolts at different positions, further includes: The total dynamic load is divided into a pneumatic load and a mechanical load; the mechanical load includes one or more of the gearbox torque, the inertia load of the rotating parts, the gravity load and the friction loss; Mechanical load ; Set up the aerodynamic load distribution matrix and calibrate it through wind tunnel experiments, including: Apply stepped wind speed in the wind tunnel and measure the strain of each bolt and the aerodynamic load vector in the global coordinate system ; Aerodynamic load distribution matrix ; Set up the mechanical load distribution matrix and calibrate it through bench tests, including: Disconnect the blades and apply stepped torque through the motor; Measure individual bolt strains and the mechanical load vector in the global coordinate system ; Mechanical load distribution matrix ; The aerodynamic load distribution matrix and the aerodynamic load vector in the global coordinate system are used to calculate the axial tensile force, shear force and torque of the bolt, and then the synthetic stress under the aerodynamic load is obtained; if the synthetic stress under the aerodynamic load exceeds the preset aerodynamic load synthetic stress threshold, it is determined that the bolt stress abnormality is caused by the aerodynamic load; The mechanical load distribution matrix and the mechanical load vector in the global coordinate system are used to calculate the axial tensile force, shear force and torque of the bolt, and then the synthetic stress under the mechanical load is obtained; if the synthetic stress under the mechanical load exceeds the preset mechanical load synthetic stress threshold, it is judged that the bolt stress abnormality is caused by the mechanical load.

6. A wind power bolt real-time stress monitoring method according to claim 5, characterized in that: The monitoring of the stress of the bolts at different positions to determine whether the bolts are faulty also includes: Construct a state-space model to capture the interaction between mechanical loads and aerodynamic loads and identify the stress distribution of the bolts. The specific steps include: Construct a state-space model, including: Obtain the deformation of the bolt connection , i.e. the rigidity of the bolt; Get the vibration velocity of the bolt , and the bolt strain ; Constructing the state vector ; Equation of state ;in, Represent the state transfer matrix, input matrix and observation matrix respectively; Input Vector ,in, Represents wind speed, blade surface Pressure and gearbox torque at each position; By using the system identification method N4SID, the matrix Parameters; through The off-diagonal elements in reflect the dynamic coupling between state vectors; The elements in reflect the influence of the elements in the input vector on the elements in the state vector; Identify stress distribution in bolts, including: The state vector is associated with the bolt stress through the observation matrix; if only the bolt stress is concerned, then , represents the identity matrix; Using the model after system identification method, input real-time load data to predict bolt stress: ,in Represent the state vector at the start time and The input vector at time instant; Get the measured stress of the bolt and calculate the difference between the predicted bolt stress and the measured stress, i.e. the stress residual ; Calculate the correlation coefficient between the stress residual of each bolt and the corresponding mechanical load distribution matrix and aerodynamic load distribution matrix, specifically: Aerodynamic load correlation coefficient , mechanical load correlation coefficient ; if It is judged that the failure of the bolt comes from the aerodynamic load; if It is judged that the failure of the bolt comes from the mechanical load; Obtain real-time wind speed data and air pressure at multiple locations on the blade surface to determine whether the air pressure matches, so as to determine whether the variables related to the aerodynamic load in the input vector are normal; measure the gearbox vibration data, extract the spectrum data when the gears are meshing, and detect whether the frequency assignment increases suddenly when the gears are meshing, so as to determine whether the variables related to the mechanical load in the input vector are normal; If the variables related to the mechanical load and the aerodynamic load are normal, but the residuals are continuously abnormal, it is judged that the mechanical load distribution matrix and the aerodynamic load distribution matrix are invalid, and the mechanical load distribution matrix and the aerodynamic load distribution matrix need to be recalibrated.

7. A wind power bolt real-time stress monitoring method according to claim 6, characterized in that: The bolt replacement process is performed to analyze and evaluate the effect of the treated bolts on the stress of other bolts, including: Obtain the bolt data after fault recovery, including the new preload and new stiffness of the bolt; New stiffness of the bolt ;in, Respectively represent the new preload and bolt elongation; Update bolts Bolt The coupling coefficient of the new coupling coefficient matrix is ​​obtained ;in, Indicates the bolts before replacement Axial stiffness; Get the historical preload force of the failed bolt from the database , update the dynamic total load distribution matrix, and obtain the updated dynamic total load distribution matrix ; use and Calculate the new axial tension, shear force and torque to obtain the new stress matrix ; Obtaining tensile stress using stress matrix calculation , shear stress and bending stress ;in, Indicates the distance from the neutral axis of the bolt section to the outermost edge. If the cross section of the bolt is circular, then , Indicates the diameter of the cross section; The resultant stress is ; Resultant stress change rate ,if , then determine the update bolt Affects the stress safety of other bolts.

8. A wind power bolt real-time stress monitoring method according to claim 7, characterized in that: The bolt replacement process and analysis and evaluation of the effect of the treated bolts on the stress of other bolts also include: Considering the effect of the installation of new bolts on the local contact pressure distribution, the effect of the preload of the new bolts on the structural deformation of the contact part, and the change of the friction coefficient under dynamic load, specifically: Measuring bolts with hydraulic tensioners Axial stiffness ; The contact pressure distribution of the bolt connection surface is measured by an ultrasonic detector to calculate the friction coefficient ; Monitoring the preload attenuation ratio of adjacent bolts by distributed strain gauges ; Correct the coupling coefficient and obtain the corrected new coupling coefficient ,in, Indicates bolt Belongs to the adjacent bolt set ; Respectively indicate replacement of front and rear bolts The friction coefficient of Bolt Apply step loads and record the measured strain responses; Building a nonlinear regression model ,in, represents the measured strain response, that is, the measured stress matrix , represents the step load vector; represents the coefficient matrix; , ; The output of the nonlinear regression model is predicted , combined with Calculate and obtain the corresponding tensile stress, shear stress and bending stress, and then obtain the corresponding synthetic stress change rate to determine the impact of the installation of the new bolt on the stress safety of other bolts.

9. A wind power bolt real-time stress monitoring system, comprising: A measurement module is used to collect the rotation speed of the motor main shaft, the air pressure at multiple locations on the blade surface, and the motor output voltage and current; The monitoring module is used to analyze and process the data input by the measurement module, monitor the stress of different screws, determine whether different bolts are faulty, and output the judgment results; A communication module is used to connect to a host computer and transmit the judgment result to the host computer; It is characterized in that the system is used to execute a wind power bolt real-time stress monitoring method as described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement a wind power bolt real-time stress monitoring method as described in any one of claims 1 to 8.

Images