Dynamic load monitoring and overload early warning method for in-service wind power blade

Abstract

The invention discloses an in-service wind power blade load dynamic monitoring and overload early warning method, which comprises the steps of analyzing blade section stress under different wind conditions, and identifying a maximum dangerous section by taking a maximum equivalent stress criterion as a failure judgment basis as an optimal monitoring position of a strain sensor; determining an optimal strain sensor installation number according to a finite element circumferential strain deformation result at the maximum dangerous section and a sensitive grid effective monitoring range of the strain sensor; establishing a resistance or voltage and load measurement model to realize dynamic load monitoring of the in-service blade; and the blade overload judgment and early warning are realized by comparing the delivery design load of the blade with the load data obtained by monitoring. According to the method, dynamic load monitoring and overload early warning of the in-service wind power blade can be accurately achieved by installing the stress sensor at the maximum dangerous section, wind power output is further optimized and controlled by synchronously increasing monitoring of wind parameters with large influences such as wind shear and turbulence, and the anti-fatigue performance of the in-service wind turbine generator blade is improved.

Description

Technical Field

[0001] This invention belongs to the field of wind turbine blade condition monitoring technology, and more specifically relates to a method for dynamic monitoring of load and overload early warning of in-service wind turbine blades. Background Technology

[0002] Wind turbines frequently experience blade failures such as leading-edge corrosion, tip detachment, icing corrosion, and adhesive bonding cracking under the combined effects of complex random loads and outdoor environments. Failure to detect and assess the type and severity of these failures early can lead to unplanned downtime and excessive maintenance, and may even result in catastrophic breakage accidents under specific operating conditions. Furthermore, the installation and commissioning of new blades and the secondary installation of upgraded blades require dynamic balancing tests, i.e., testing the consistency of loads on different blades within the same turbine unit, necessitating a precise blade load monitoring system. Additionally, the control system of in-service wind turbines can accurately execute actions such as wind control, pitch adjustment, shutdown, and startup based on real-time blade loads, providing crucial foundational data for improving wind turbine power generation efficiency. Therefore, real-time monitoring and identification of in-service blade loads are essential for the safe operation and cost reduction / efficiency improvement of wind turbines. Currently, large wind turbine blades exceed 100 meters in length and have hundreds of square meters of surface area. Considering the economic, technical, and reliable nature of monitoring systems, commercial blade load monitoring systems often use strain sensors for real-time monitoring. However, in actual wind farm applications, the following technical challenges remain: (1) During the 20-year design service life, the blades of wind turbines will experience 108 to 109 cycles of load, which is much greater than the number of cycles of load that vehicles, helicopters and other mechanical equipment can withstand. This makes the fatigue load characteristics of the blades extremely complex under the combined effects of gravity load, aerodynamic load and random load, and the arrangement of sensors lacks theoretical basis. (2) The airfoil structure, ply design, material properties and manufacturing process of blades designed by different manufacturers are quite different. Their fatigue resistance is more uncertain under varying wind conditions, making it impossible to achieve accurate overload warning. This results in a very high false alarm rate for overload monitoring of in-service blades and a lack of universality of blade load monitoring systems. (3) When the same type of blade is installed in different wind fields, its terrain roughness, wind profile, wind shear and turbulence may differ greatly from the original design, resulting in a large difference in the wind speed distribution along the height near the monitored wind turbine, which leads to a huge difference between the actual monitored dynamic load and the original design load, making it difficult to judge the degree of influence of wind condition parameters on the safe operation of the blade in service. Summary of the Invention

[0003] In view of this, the present invention addresses the shortcomings of the prior art by providing a method for dynamic monitoring and overload warning of loads on in-service wind turbine blades. By installing stress sensors at the maximum critical section, the method can accurately achieve dynamic monitoring and overload warning of loads on in-service wind turbine blades. Furthermore, by simultaneously increasing the monitoring of wind parameters with significant impacts such as wind shear and turbulence, the method can further optimize the control of wind power output and improve the fatigue resistance of in-service wind turbine blades.

[0004] To achieve the above objectives, the present invention adopts the following technical solution: A method for dynamic monitoring and overload early warning of loads on in-service wind turbine blades includes the following steps: Step S1: Strain sensor monitoring layout; Based on the three-dimensional model of the in-service blade and wind environmental load parameters, establish a Fluent finite element simulation calculation model, analyze the stress on the blade section under different wind conditions, and use the maximum equivalent stress criterion (VonMises failure criterion) as the basis for failure judgment to identify the maximum critical section as the optimal monitoring location for strain sensors; Determine the optimal number of strain sensors to be installed based on the finite element circumferential strain deformation results at the maximum critical section and the effective monitoring range of the sensitive grid of the strain sensor. Step S2: In-service blade fatigue load monitoring; Strain sensors are attached to the optimal position on the blade surface to obtain strain resistance or voltage data at the measuring points during the operation of the blade. By calculating the equivalent fatigue load of the wind turbine blade, load data at different cross-sectional positions of the wind turbine blade are obtained. The stress-load relationship is established based on the mechanical properties of the material, and then a measurement model of resistance or voltage and load is established to realize dynamic load monitoring of in-service blades. Step S3: Load monitoring data analysis and evaluation; First, the data collected by data acquisition devices such as strain sensors, SCADA systems, and lidar systems are analyzed to determine data changes and acquisition times, thereby determining the blade orientation, or judging the blade orientation and pitch angle changes during the slow rotation of the blade based on the time sequence; Then, the spectral characteristics of the response are extracted from the strain measurement results, and the natural frequencies and modal characteristics of the blade are calculated by combining finite element simulation to determine the load trend of the blade in flapping and oscillating states; Finally, by comparing the load data obtained from monitoring with the design load of the blade at the factory, blade overload judgment and early warning are achieved.

[0005] Furthermore, in step S1, the method for determining the optimal monitoring location of the strain sensor and the optimal number of strain sensors installed includes: Step S11: Using professional finite element analysis software, construct a high-precision three-dimensional finite element model of the blade according to its actual size, material properties, and internal structure. Step S12: After the three-dimensional finite element model of the blade is constructed, parameters under different operating conditions are obtained from the database of actual wind turbine operation. These parameters are then input into the three-dimensional finite element model of the blade to simulate the stress on the blade under various operating conditions. Step S13: Analyze the simulation results. By examining the overall stress cloud diagram of the blade, identify the areas on the blade with the most severe stress concentration and the largest stress value, and determine them as the areas with the largest critical sections. Step S14: For the area with the maximum critical section, based on the finite element circumferential strain deformation results at the maximum critical section and the effective monitoring range of the strain sensor, determine the optimal monitoring position of the strain sensor and the optimal number of strain sensors to be installed.

[0006] Furthermore, in step S12, the parameters under different operating conditions include: wind speed range from 3m / s to 25m / s, wind direction change between 0° and 360°, blade pitch angle adjustment from 0° to 90°, and different rotational speeds, etc.

[0007] Furthermore, in step S2, after realizing the dynamic load monitoring of the in-service blades, the accuracy of the equivalent fatigue load calculation results is verified by comparing the monitored loads of the same blades and adjacent blades with the calculated equivalent fatigue loads; the blade load is calibrated using the bending moment caused by gravity.

[0008] Furthermore, in step S2, the method for monitoring the dynamic load on in-service blades includes: Step S21: On the established three-dimensional finite element model of the blade, start the modal analysis function and calculate the natural frequency and corresponding mode shape of the blade through software. Step S22: Verify the calculated natural frequency data by actually measuring the natural frequency of the in-service blade using vibration testing equipment; if there is a large deviation between the calculated value and the theoretical or measured value, correct the finite element model until the model calculation result matches the actual situation; Step S23: During the test, the wind turbine should be in normal operating condition, suitable for comparing fatigue loads. Since the blade material is an epoxy resin-based laminate, the slope m of the material's SN curve is calculated according to the GL2010 wind turbine blade certification standard guideline, and the equivalent fatigue load is calculated using this slope value. Step S24: Calculate the equivalent design fatigue load at 1Hz corresponding to the blade's factory design fatigue load parameters, using the following formula: (1) Among them, R eq For the equivalent design fatigue load at 1Hz, R df For designing fatigue loads, N is the number of cycles, and m is the slope of the material's SN curve; Step S25: Calculate the design fatigue load for different cross sections of the wind turbine blade to obtain the equivalent design fatigue load at 1Hz for different cross section positions of the wind turbine blade. Step S26: Compare the monitoring load of the same blade and adjacent blades with the calculated equivalent fatigue load, evaluate whether the equivalent fatigue load calculation result will be affected by the wind field monitoring environment, and verify the accuracy of the equivalent fatigue load calculation result. Step S27: Determine the blade orientation by judging the changes in the collected data and the time of collection, or judge the changes in blade orientation and pitch angle during the slow rotation of the blade based on the time sequence. Step S28: Since the blade is affected by gravity during rotation, the bending moment in the flapping and oscillation directions at the test section can be obtained based on the bending moment caused by gravity. Step S29: Based on the analysis results of the optimal monitoring position of the sensor, strain gauges are attached to the most dangerous area of ​​the blade to obtain the strain signal of the blade under load. Step S210: Based on the linear relationship between strain and load at the measuring point, the strain of the blade during operation is obtained by measuring the resistance change of the strain gauge. Then, the stress and load borne by the blade are calculated based on the mechanical properties of the material, thereby realizing the dynamic load monitoring of the blade in service.

[0009] Furthermore, for comparative analysis with design and simulation loads, the ten-minute time series was calculated using an equivalent fatigue load of 1Hz, as shown in the formula: (2) in, For a 10-minute time series 1Hz equivalent load, R i For the i-th interval of the load spectrum, n i The corresponding range value is the number of iterations, m is the slope of the material's SN curve, and N is the number of cycles.

[0010] Furthermore, in step S3, the specific methods for realizing blade overload judgment and early warning include: Step S31: Select wind environment test data in the sequence of time when the unit is operating, the data is within the normal range, and the wind speed measurement is normal; Step S32: Analyze the data file duration of ten minutes according to IEC61400-13 requirements, use lidar to collect wind speed and direction information at 2.5 times the rotor diameter, and divide the load data into measurement files of ten minutes each according to the average wind measurement data exported from lidar. Step S33: Based on the strain results obtained from the actual monitoring of the blade, perform time-series spectrum analysis on the strain data to extract the characteristic frequencies of the blade. Step S34: Based on the spectral response of the blade's characteristic frequency, compare the modal response calculation results to determine the blade's vibration mode characteristics under the current wind load, namely the flapping direction and the oscillation direction. Step S35: Based on the strain monitoring results and blade vibration mode comparison results, obtain the load values ​​and trends of the blade under the corresponding vibration modes in the flapping and swaying directions under wind conditions. Step S36: The strain sensor on the blade collects data such as stress, strain, deformation and load of the blade in real time and transmits it to the data processing center. In step S37, the data processing center compares the real-time collected data with the fatigue load designed for the blade at the factory. If the monitored data exceeds any fatigue overload threshold, it indicates that the current load of the blade is at risk of overload and an early warning is required.

[0011] The beneficial effects of this invention are as follows: (1) The theoretical basis and arrangement method for optimizing the location and number of sensors proposed can effectively improve the accuracy of dynamic load monitoring under complex wind field conditions, and reduce costs and increase efficiency.

[0012] (2) The established dynamic fatigue load sensor measurement model and sensor parameter calibration method are highly universal and can meet the needs of different blades, different manufacturers and different wind conditions, ensuring the reliability of overload warning and being of great significance for early detection of faults and safe service.

[0013] (3) The analysis results of the influence of wind parameters under complex working conditions provide new monitoring basis for wind farms and provide solutions for the large differences between actual wind farm conditions and original design. It can more accurately monitor and identify the dynamic load of blades in real time, effectively prevent overload operation, optimize the wind power output of the control system, and improve the fatigue resistance of the blades of the wind turbine units in service. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0015] Figure 1 It is a roadmap for load dynamic monitoring and overload early warning methods.

[0016] Figure 2 This is a diagram showing the distribution of bending moment caused by gravity along the blade span when the blade is in a horizontal state.

[0017] Figure 3 This is a schematic diagram of the fluid finite element calculation model for wind turbine blades.

[0018] Figure 4 This is a schematic diagram of the overall strength calculation results for the blade.

[0019] Figure 5 This is a schematic diagram of the blade modal calculation results.

[0020] Figure 6 This is a schematic diagram of the collected blade strain signals.

[0021] Figure 7 It is the wind profile in complex terrain.

[0022] Figure 8 This is a graph showing the results of the spectral analysis. Detailed Implementation

[0023] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0024] To address the problems of inaccurate dynamic load monitoring and high false alarm rates under complex wind farm conditions caused by unreasonable sensor layout in existing wind turbine blade load measurement methods, this invention proposes a method for dynamic load monitoring and overload early warning of in-service wind turbine blades. Based on the sensor layout theory of the maximum equivalent stress criterion, the optimal installation position and number of sensors are determined by the maximum cross-section, enabling rapid and accurate measurement of dynamic fatigue loads under complex multi-load conditions. The method compares design loads, test loads, and simulated loads to resolve uncertainties such as large differences between actual operating conditions and original designs, and variable wind conditions. By comparing equivalent fatigue loads under different operating parameters such as terrain roughness, wind profile, wind shear, and turbulence intensity in actual wind farms, the method effectively distinguishes the degree of influence of wind parameters on the safe operation of in-service blades. The method proposed in this invention can improve the accuracy of dynamic load monitoring of in-service wind turbine blades and reduce the false alarm rate of overload.

[0025] like Figure 1 As shown, this embodiment provides a method for dynamic monitoring and overload early warning of loads on in-service wind turbine blades, including the following steps: Step 1: Strain sensor monitoring layout, including the following methods: Step 1.1: Determine the optimal monitoring location and number of sensors using the maximum equivalent stress criterion (Von Mises failure criterion). The maximum equivalent stress criterion states that under certain deformation conditions, when the equivalent stress at a point within a stressed object reaches the yield point of the material, that point begins to enter a plastic state. The equivalent stress is calculated as shown in equation (1). Therefore, the area of ​​stress concentration and maximum deformation on the blade is the most dangerous cross-sectional area where the blade is most prone to failure, which is the optimal monitoring location.

[0026] (1) in, Representing the materials in , and Principal stresses in the direction. These stresses are normal stresses acting in three mutually perpendicular directions. Representing the materials in , and Shear stresses on a plane. These stresses describe the material's ability to undergo shear deformation in different planes. This is the yield point of the material, i.e., the stress value at which the material begins to undergo plastic deformation. This is the stress value at which the material reaches its yield point in a uniaxial tensile test. Equivalent stress is an equivalent value that combines the effects of normal stress and shear stress, and is used to evaluate the yielding behavior of materials under complex stress states.

[0027] Step 1.2: Obtain the design parameters of the in-service blade structure and materials. Obtain wind environmental load parameters such as wind speed and wind direction through the SCADA system of the in-service wind turbine.

[0028] Step 1.3: Establish a three-dimensional Fluent finite element simulation model of the blade. Through simulation of various working conditions, comprehensively analyze the stress distribution and deformation of the blade.

[0029] Step 1.4: The area with the most severe stress concentration and the greatest deformation on the blade is the maximum dangerous section, which needs to be monitored. This is the optimal location for placing sensors, and the specific placement locations are shown in Table 1.

[0030] Table 1. Location of Maximum Cross-Sectional Strain Sensors Step 1.5: Determine the optimal number of sensors to be installed based on the finite element circumferential strain deformation results at the maximum critical section and the effective monitoring range of the strain sensor's sensitive grid.

[0031] Step 2: Monitoring fatigue load on in-service blades, including the following methods: Step 2.1: Attach the strain sensor at the optimal position on the blade surface. The resistance of the strain sensor changes with the deformation of the blade. The relationship is (2). Obtain the strain resistance (voltage) data at the measuring point of the blade during operation.

[0032] (2) In the formula, ΔR is the change in resistance, and R is the resistance. Let K be the stress, and K be the strain sensitivity coefficient of the strain sensor.

[0033] Step 2.2: Obtain the load calculation results at different cross-sectional positions of the wind turbine blade by calculating the equivalent fatigue load of the wind turbine blade.

[0034] Step 2.3: Establish a resistance (voltage)-stress-load measurement model to monitor the load on in-service blades.

[0035] Step 2.4: Verify the accuracy of the equivalent fatigue load calculation results by comparing the monitoring load of the same blade and adjacent blades with the calculated equivalent fatigue load.

[0036] Step 2.5: Calibrate the blade load using the bending moment caused by gravity. The bending moment caused by gravity under horizontal blade conditions is distributed along the blade spanwise as follows: Figure 2 As shown.

[0037] Step 3: Load monitoring data analysis and evaluation, including the following methods: Step 3.1: Data Time Series Analysis. First, by analyzing the data changes and acquisition times collected by devices such as strain sensors, SCADA systems, and lidar systems, the azimuth of the measured blades is determined, or the changes in blade azimuth and pitch angle during the slow rotation of the blades are determined based on the time series. Strain data is continuously acquired, with one file saved every hour at a sampling frequency of 64.516Hz. The amount of unit control information acquired is shown in Table 2. The real-time wind speed data acquisition cycle of the lidar is 4 seconds.

[0038] Table 2 Control Information Collection Step 3.2: Frequency Domain Analysis of Data. The spectral characteristics of the response are extracted from the strain measurement results. Combined with finite element simulation, the natural frequencies and modal characteristics of the blade are calculated to determine the load trends of the blade in flapping and oscillating states.

[0039] Step 3.3: By comparing the load data obtained from monitoring with the design load of the blade at the factory, overload warning of the blade can be achieved.

[0040] Further optimization of the technical solution involves determining the optimal installation location and number of strain sensors in step 1. This method enables accurate acquisition of blade loads, specifically including: Step 1.1.1: Using professional finite element analysis software, construct a high-precision three-dimensional finite element model of the blade according to its actual dimensions, material properties, and internal structure. Figure 3 As shown.

[0041] Step 1.1.2: After the model is built, parameters for different operating conditions are obtained from the database of actual wind turbine operation, including wind speed range from 3m / s to 25m / s, wind direction variation between 0° and 360°, blade pitch angle adjustment from 0° to 90°, and different rotational speeds. These parameters are then input into the finite element model to simulate the stress on the blades under various operating conditions, such as... Figure 4 As shown.

[0042] Step 1.1.3: Analyze the simulation results. By examining the overall stress cloud diagram of the blade, identify the areas on the blade with the most severe stress concentration and the highest stress value. The most dangerous area is the optimal installation location for the sensor.

[0043] Step 1.1.4: Determine the specific number of sensors based on the strain results of the maximum danger zone and the effective monitoring range of the strain sensor's sensitivity.

[0044] Step 2 describes a method for monitoring the fatigue load of wind turbine blades. This method calculates the fatigue load of wind turbine blades under different wind environments using equivalent fatigue loads, establishes the conversion relationship between strain signals and loads, and enables monitoring of the blade's load condition based on strain signals. Specifically, this includes: Step 2.1.1: On the existing finite element model of the blade, start the modal analysis function. Some results are as follows: Figure 5 As shown, the natural frequencies corresponding to the first-order flapping, first-order oscillation, second-order flapping, and second-order oscillation modes of the blade are calculated by software.

[0045] Step 2.1.2: Verify the calculated natural frequency data by actually measuring the natural frequencies of the in-service blades using vibration testing equipment. If there is a large deviation between the calculated value and the theoretical or measured value, correct the finite element model until the model calculation results match the actual situation.

[0046] Step 2.1.3: During the test, the wind turbine should be in normal operating condition, suitable for comparing fatigue loads. Since the blade material is an epoxy resin-based laminate, the slope value m of the material's SN curve is calculated according to the GL2010 wind turbine blade certification standard guideline, and the equivalent fatigue load is calculated using this slope value.

[0047] Step 2.1.4: To make comparisons under the same conditions, calculate the design equivalent fatigue load at 1Hz. The calculation formula is as follows: (3) Among them, R eq For the equivalent design fatigue load at 1Hz, R df To design the fatigue load, N is the number of cycles, and m is the slope of the material's SN curve. Here, m = 10 is taken.

[0048] Step 2.1.5: Use equation (3) to perform equivalent calculations on the design equivalent fatigue load of the wind turbine blade and obtain the equivalent fatigue load at different cross-sectional positions of the wind turbine blade.

[0049] Step 2.1.6: Compare the monitoring load of the same blade and adjacent blades with the calculated equivalent fatigue load to assess whether the equivalent fatigue load calculation results will be affected by the wind field monitoring environment interference, and verify the accuracy of the equivalent fatigue load calculation results.

[0050] Step 2.1.7: Determine the blade orientation by analyzing the changes in the collected data and the timing of the data collection, or determine the changes in blade orientation and pitch angle during the slow rotation of the blade based on the time sequence.

[0051] Step 2.1.8: Since the blade is affected by gravity during rotation, the bending moment in the flapping and oscillation directions at the test section can be obtained based on the bending moment caused by gravity.

[0052] Step 2.1.9: Based on the analysis results of the optimal monitoring position of the sensor, strain gauges are attached to the most dangerous area of ​​the blade to obtain the strain signal of the blade under load. Some data are shown below. Figure 6 As shown.

[0053] Step 2.1.10: Based on the linear relationship between strain and load at the measuring point, as shown in formulas (4) and (5), the strain of the blade during operation is obtained by measuring the resistance change of the strain gauge. Then, the stress and load borne by the blade are calculated based on the mechanical properties of the material, so as to realize the load monitoring of the blade in service.

[0054] (4) (5) In the formula, M F For the blade flapping bending moment, M E β is the blade oscillation bending moment, β is the blade pitch angle, and k1 and k2 are sensor calibration factors. Their values ​​depend on the characteristics of the blade material and shape. Different blades in the same unit also need to be calibrated individually. , , , These represent the strain values ​​at the pressure surface, suction surface, trailing edge, and leading edge of the measuring point, respectively.

[0055] Step 2.1.11: To compare and analyze the loads with the design and simulation loads, the ten-minute time series of the test load data is calculated as an equivalent fatigue load of 1Hz using the following formula: (6) in, For a 10-minute time series 1Hz equivalent load, R i For the i-th interval of the load spectrum, n i The corresponding range value is the number of iterations, m is the slope of the material's SN curve, here we take m=10, and N is the number of cycles.

[0056] Step 3 involves analyzing and evaluating the wind turbine blade load test data. Based on the strain results obtained from actual blade monitoring, time-series spectral analysis is performed on the strain data to extract the blade's characteristic frequencies. By comparing the spectral response of these characteristic frequencies with the modal response calculation results, the vibration mode characteristics of the blade under the current wind load can be determined, i.e., the flapping and swaying directions. Based on the strain results, the magnitude and trend of the load under the corresponding vibration mode can be calculated. Comparing this with the fatigue load designed for the blade at the factory allows for risk warnings regarding the load on in-service blades. Specific methods include: Step 3.1.1: Select wind environment test data based on the time sequence when the unit is operating normally, the data is within the normal range, and the wind speed measurement is normal. The collected partial wind condition distribution is shown below. Figure 7 As shown, Figure 7 (a) shows the wind profile under complex terrain during the day. Figure 7 (b) is the wind profile under complex terrain at night.

[0057] Step 3.1.2: According to IEC61400-13 requirements, use lidar to collect wind speed and direction information at a distance of 2.5 times the rotor diameter. The load data is divided into 10-minute measurement files based on the average wind measurement data exported from the lidar.

[0058] Step 3.1.3: Based on the strain results obtained from actual blade monitoring, perform spectral analysis on the strain time series data. Some results are shown below. Figure 8 As shown, by comparing the theoretical frequency values ​​with those analyzed by the finite element method, the first-order flapping, first-order oscillation, second-order flapping, and second-order oscillation of the blade can be identified.

[0059] Step 3.1.4: Based on the spectral response of the characteristic frequency, compare the modal response calculation results to determine the vibration characteristics of the blade under the current wind load, namely the flapping direction and the swaying direction.

[0060] Step 3.1.5: Based on the strain monitoring results and blade vibration mode comparison results, obtain the load values ​​and trends of the blade under the corresponding vibration modes in the flapping and swaying directions under wind conditions.

[0061] Step 3.1.6: The strain sensors on the blade collect data such as stress, strain, deformation and load in real time and transmit them to the data processing center.

[0062] Step 3.1.7: The data processing center compares the real-time collected data with the fatigue load designed for the blade at the factory. If the monitored data exceeds any fatigue overload threshold, it indicates that the current load of the blade is at risk of overload and an early warning is required.

[0063] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for dynamic monitoring and overload early warning of loads on in-service wind turbine blades, characterized in that, Includes the following steps: Step S1: Strain sensor monitoring layout; Based on the three-dimensional model of the in-service blade and the wind environment load parameters, establish a Fluent finite element simulation calculation model, analyze the stress of the blade section under different wind conditions, and use the maximum equivalent stress criterion as the failure judgment basis to identify the maximum dangerous section as the optimal monitoring position of the strain sensor. The optimal number of strain sensors to be installed is determined based on the finite element circumferential strain deformation results at the maximum critical section and the effective monitoring range of the strain sensor's sensitive grid. Step S2: Monitoring fatigue load on in-service blades; Strain sensors are attached to the optimal position on the blade surface to acquire strain resistance or voltage data at the measuring points during the operation of the blade. By calculating the equivalent fatigue load of the wind turbine blade, load data at different cross-sectional positions of the wind turbine blade are obtained. The stress-load relationship is established based on the mechanical properties of the material, and then a measurement model of resistance or voltage and load is established to realize dynamic load monitoring of the in-service blade. Step S3: Load monitoring data analysis and evaluation; First, the data collected by the strain sensor, SCADA system and lidar system are analyzed to determine the data changes and acquisition time, and the position of the measured blade is determined, or the changes in blade position and pitch angle during the slow rotation of the blade are determined according to the time sequence. Then, the spectral characteristics of the response are extracted from the strain measurement results, and the natural frequency and modal characteristics of the blade are calculated by combining finite element simulation to determine the load trend of the blade in flapping and oscillating states. Finally, by comparing the load data obtained from monitoring with the design load of the blade at the factory, the blade overload judgment and early warning are realized.

2. The method for dynamic monitoring and overload early warning of load on in-service wind turbine blades according to claim 1, characterized in that, In step S1, the methods for determining the optimal monitoring location of the strain sensor and the optimal number of strain sensors to be installed include: Step S11: Using professional finite element analysis software, construct a high-precision three-dimensional finite element model of the blade according to its actual size, material properties, and internal structure. Step S12: After the three-dimensional finite element model of the blade is constructed, parameters under different operating conditions are obtained from the database of actual wind turbine operation. These parameters are then input into the three-dimensional finite element model of the blade to simulate the stress on the blade under various operating conditions. Step S13: Analyze the simulation results. By examining the overall stress cloud diagram of the blade, identify the areas on the blade with the most severe stress concentration and the largest stress value, and determine them as the areas with the largest critical sections. Step S14: For the area with the maximum critical section, based on the finite element circumferential strain deformation results at the maximum critical section and the effective monitoring range of the strain sensor, determine the optimal monitoring position of the strain sensor and the optimal number of strain sensors to be installed.

3. The method for dynamic monitoring and overload early warning of load on in-service wind turbine blades according to claim 2, characterized in that, In step S12, the parameters under different operating conditions include: wind speed range from 3m / s to 25m / s, wind direction change between 0° and 360°, blade pitch angle adjustment from 0° to 90°, and different rotational speeds.

4. The method for dynamic monitoring and overload early warning of load on in-service wind turbine blades according to claim 1, characterized in that, In step S2, after the dynamic load monitoring of the in-service blades is realized, the accuracy of the equivalent fatigue load calculation results is verified by comparing the monitored loads of the same blades and adjacent blades with the calculated equivalent fatigue loads; the blade load is calibrated using the bending moment caused by gravity.

5. The method for dynamic monitoring and overload early warning of load on in-service wind turbine blades according to claim 4, characterized in that, In step S2, the method for monitoring the dynamic load on in-service blades includes: Step S21: On the established three-dimensional finite element model of the blade, start the modal analysis function and calculate the natural frequency and corresponding mode shape of the blade through software. Step S22: Verify the calculated natural frequency data by actually measuring the natural frequency of the in-service blade using vibration testing equipment; if there is a large deviation between the calculated value and the theoretical or measured value, correct the finite element model until the model calculation result matches the actual situation; Step S23: During the test, the wind turbine should be in normal operation. According to the GL2010 wind turbine blade certification standard guide, calculate the slope m of the material SN curve, and use this slope value to calculate the equivalent fatigue load. Step S24: Calculate the equivalent design fatigue load at 1Hz corresponding to the blade's factory design fatigue load parameters, using the following formula: （1） Among them, R eq For the equivalent design fatigue load at 1Hz, R df For designing fatigue loads, N is the number of cycles, and m is the slope of the material's SN curve; Step S25: Calculate the design fatigue load for different cross sections of the wind turbine blade to obtain the equivalent design fatigue load at 1Hz for different cross section positions of the wind turbine blade. Step S26: Compare the monitoring load of the same blade and adjacent blades with the calculated equivalent fatigue load, evaluate whether the equivalent fatigue load calculation result will be affected by the wind field monitoring environment, and verify the accuracy of the equivalent fatigue load calculation result. Step S27: Determine the blade orientation by judging the changes in the collected data and the time of collection, or judge the changes in blade orientation and pitch angle during the slow rotation of the blade based on the time sequence. Step S28: Obtain the bending moment in the swinging and oscillating directions at the test section based on the bending moment caused by gravity; Step S29: Based on the analysis results of the optimal monitoring position of the sensor, strain gauges are attached to the most dangerous area of ​​the blade to obtain the strain signal of the blade under load. Step S210: Based on the linear relationship between strain and load at the measuring point, the strain of the blade during operation is obtained by measuring the resistance change of the strain gauge. Then, the stress and load borne by the blade are calculated based on the mechanical properties of the material, thereby realizing the dynamic load monitoring of the blade in service.

6. The method for dynamic monitoring and overload early warning of load on in-service wind turbine blades according to claim 5, characterized in that, To compare with the design and simulation loads, the ten-minute time series was calculated using a 1Hz equivalent fatigue load, as shown in the formula: （2） in, For a 10-minute time series 1Hz equivalent load, R i For the i-th interval of the load spectrum, n i The corresponding range value is the number of iterations, m is the slope of the material's SN curve, and N is the number of cycles.

7. The method for dynamic monitoring and overload early warning of load on in-service wind turbine blades according to claim 1, characterized in that, In step S3, the specific methods for realizing blade overload judgment and early warning include: Step S31: Select wind environment test data in the sequence of time when the unit is operating, the data is within the normal range, and the wind speed measurement is normal; Step S32: Analyze the data file duration of ten minutes according to IEC61400-13 requirements, use lidar to collect wind speed and direction information at 2.5 times the rotor diameter, and divide the load data into measurement files of ten minutes each according to the average wind measurement data exported from lidar. Step S33: Based on the strain results obtained from the actual monitoring of the blade, perform time-series spectrum analysis on the strain data to extract the characteristic frequencies of the blade. Step S34: Based on the spectral response of the characteristic frequency of the blade, compare the modal response calculation results to determine the vibration mode characteristics of the blade under the current wind load, namely the flapping direction and the oscillation direction. Step S35: Based on the strain monitoring results and blade vibration mode comparison results, obtain the load values ​​and trends of the blade under the corresponding vibration modes in the flapping and swaying directions under wind conditions. Step S36: The strain sensor on the blade collects data such as stress, strain, deformation and load of the blade in real time and transmits it to the data processing center. In step S37, the data processing center compares the real-time collected data with the fatigue load designed for the blade at the factory. If the monitored data exceeds any fatigue overload threshold, it indicates that the current load of the blade is at risk of overload and an early warning is required.

Images