A wind turbine blade health monitoring device and monitoring method

Abstract

The invention discloses a wind turbine blade health monitoring device and a monitoring method. The device supports a sound level meter and a protective cover through radial brackets and circumferential brackets, and measures audio information in multiple directions at the same time. The method analyzes noise characteristic parameters in the frequency domain space, removes the interference of environmental noise on the unit noise, extracts blade noise information by using the wind wheel rotation speed, and locates the spatial distribution law of the blade noise characteristic parameters by using the unit yaw angle. The frequency domain energy amplitude under the discrete parameter combination is extracted by discrete environmental temperature, wind wheel rotation speed, and relative angle direction parameters, and the measurement error is corrected. The discrete parameter value domain integral is used to define the change amount of the blade noise characteristic distribution state result. The device and method realize the monitoring and judgment of the blade operation health state by collecting, analyzing and evaluating the noise data during the operation of the wind turbine, and provide technical support for the wind turbine blade health monitoring work.

Description

Technical Field

[0001] The invention belongs to the field of wind turbine blade health monitoring, and relates to a wind turbine blade health monitoring device and a monitoring method. Background Art

[0002] In order to ensure the continuous and efficient operation of wind farms, the technical level of health monitoring of wind turbines is also constantly improving. Due to the harsh working environment, the rotor blades of wind turbines are one of the large components that are more prone to performance degradation or even failure. Therefore, blade health monitoring has always been valued by technical personnel in the industry. After long-term operation, the blades may show the following problems: surface stains, peeling of the leading edge coating, shedding of rubber blocks in the cavity, cracking of the trailing edge adhesive, expansion of wrinkles and cracks in the inner skin ply, etc. As the blade design size becomes larger and larger, traditional health monitoring methods are difficult to achieve good detection results. The reasons are:

[0003] From the perspective of methods, (a) visual observation: the monitoring effect is often limited by the visible distance of the environment and the movement state of the blade, making it difficult to detect small damage and its degree changes in time; (b) stress monitoring: the monitoring effect is often limited by the number of sensor layout positions, instrument resolution and reliability. Due to investment and operation and maintenance costs, there is a high probability of monitoring omissions for damage that is not correctly predicted in advance; (c) vibration monitoring: in addition to facing the same problems as stress monitoring, the monitoring effect is often limited by large blade deformation and nonlinear vibration coupling, making it difficult to obtain sufficient information to accurately judge the health problems of the blade.

[0004] From the perspective of instruments, (a) traditional health monitoring methods usually install sensors directly on the research system for data collection and measurement. The operating environment of large wind turbines is complex, with many influencing factors and common system coupling responses. The signals obtained will be interfered to varying degrees, and even the performance or service life of the instrument will be reduced; (b) as the demand becomes increasingly clear, advanced technologies in other fields have also begun to be applied across industries, resulting in remote monitoring methods using video or audio. The sensor maintains a certain distance from the research system, effectively avoiding signal interference from the wind turbine itself, but cannot avoid being affected by the surrounding environment, such as rain, snow, fog, noise, etc. Instruments are also prone to failure or deviation when operating in the field for a long time. It is crucial to use technical methods such as installation and commissioning, measurement and calibration to ensure the effectiveness of the results.

[0005] In fact, for most of the current wind farms, although the SCADA system is installed, the traditional manual inspection method is still widely retained, especially in the operation health inspection of wind turbines, which relies on sound recognition and shows its intuitive and effective characteristics. The basic principle is that vibration, as an inherent property of the system, will change after a certain degree of damage or failure, and will be manifested in the form of changes in operating noise, so that early warning effects can be achieved through monitoring and evaluation. However, the shortcomings of manual inspection are also very obvious: (a) The frequency of periodic inspections and the continuous observation time are limited, so that the inspection results are significantly affected by the randomness of sampling, and it is impossible to find and feedback problems in time; (b) Human judgment is affected by psychological effects, and the reference value of the results depends on the experience and status of the inspectors, which makes it difficult to achieve objective evaluation, detail quantification, and deep learning; (c) Human auditory discrimination and judgment abilities have physiological limitations and cannot filter or capture the impact of subtle changes in on-site environmental factors.

[0006] There are technical standards for wind turbine noise measurement, see GB / T 22516 "Wind Turbine Noise Measurement Methods", but its content is mainly for the test measurement methods of related noise, and there is no reasonable solution for the problems that may be encountered in the continuous health monitoring process, such as instrument positioning calibration, data collection deviation correction, data result evaluation and judgment, etc. The lack of corresponding technical solutions may seriously affect the effect of the test implementation. Therefore, it is necessary to design and establish a wind turbine blade health monitoring platform, and seek a commonly used and effective wind turbine blade health monitoring method that is suitable for various wind farm operating environments, blade models, etc., to capture the noise changes caused by blade damage or failure, reasonably evaluate and give early warnings, shorten the safety problem discovery cycle, and reduce the unit operation and maintenance costs. Summary of the invention

[0007] The purpose of the present invention is to solve the problems in the prior art and to provide a wind turbine blade health monitoring device and a monitoring method, aiming to solve the technical problems of the accessibility limitation of the monitoring device and the effectiveness limitation of the monitoring method in the prior art of blade health monitoring.

[0008] In order to achieve the above object, the present invention adopts the following technical solutions:

[0009] The present invention proposes a wind turbine blade health monitoring device, comprising a radial support, a circumferential support, a sound level meter and a protective cover;

[0010] The radial bracket is a cantilever support structure, one end of the radial bracket is radially mounted on the outer side wall of the tower and is evenly arranged around the tower, and the installation position of the radial bracket is higher than the tower door; the sound level meter and the protective cover are mounted on the other end of the radial bracket; there are a plurality of radial brackets, sound level meters and protective covers, and the number of radial brackets, sound level meters and protective covers is equal;

[0011] The circumferential support is an annular support structure, and the circumferential support is circumferentially mounted on the radial support.

[0012] Preferably, the radial bracket is further provided with a posture adjustment device, and the radial bracket is connected to the circumferential bracket through the posture adjustment device; the radial bracket is connected to the sound level meter and the protective cover through the posture adjustment device.

[0013] Preferably, the posture adjustment device is a connecting fastener or a bolt knob.

[0014] Preferably, the circumferential supports are arranged rotationally symmetrically with respect to the axis of the tower.

[0015] Preferably, the number of the circumferential supports is several.

[0016] The present invention proposes a monitoring method for a blade health monitoring device of a wind turbine generator set, comprising the following steps:

[0017] Step 1: Installation of wind turbine blade health monitoring device:

[0018] One of the sound level meters and protective covers is located in the wake direction of the unit in the main wind direction during installation, which is defined as monitoring channel #1; the spatial arrangement of several sound level meters and protective covers is located in a plane perpendicular to the tower axis, and several sound level meters and protective covers are evenly distributed in the circumferential position with the intersection of the tower axis and the plane where the sound level meters and protective covers are located as the center of the circle; a total of N monitoring channels are used, which are arranged in the positive direction of the yaw angle and are defined as monitoring channels #2, #3, ..., #N in sequence; the normal yaw angle of the unit to the wind in the main wind direction during installation is recorded as φ0, and the absolute angle direction θ corresponding to the monitoring channel #n is defined n =φ0+2π(n-1) / N, where n=1, 2, ..., N;

[0019] Step 2: Collection of noise data during the operation of wind turbines:

[0020] It includes the debugging of the wind turbine blade health monitoring device and the collection of wind turbine operation audio data, realizing the functions of wind turbine blade health monitoring data collection, transmission, storage and call;

[0021] Step 3: Analysis of noise data during wind turbine operation:

[0022] It includes blade health monitoring data screening, effective data parameter extraction, blade noise characteristic distribution state parameter calculation and accumulation, realizing the wind turbine blade health monitoring data screening and analysis functions;

[0023] Step 4: Evaluation of noise data during wind turbine operation:

[0024] It includes the classification and correction of blade noise characteristic distribution state parameters, the construction of blade noise characteristic distribution state function, the acquisition of result change and the judgment of blade operation health status, so as to realize the wind turbine blade health monitoring data evaluation function.

[0025] Preferably, in step 2, the collection of noise data during the operation of the wind turbine generator set includes the following steps:

[0026] R1. Blade health monitoring device debugging: When the unit is in shutdown state and there is no abnormal sound in the environment, the time domain audio data of each monitoring channel is continuously collected for a collection time length U, and the collection data is transmitted and stored; the data is called for spectrum analysis to obtain the environmental noise spectrum data results of each monitoring channel {(θ n ,Ac n (f)| stop ), n=1, 2, ..., N}, where Ac n (f) is the frequency domain energy amplitude of the ambient noise of monitoring channel #n, f is the frequency, and the subscript stop indicates the shutdown state of the unit;

[0027] R2. Wind turbine operation audio data collection: When the unit is in operation, the time domain audio data of each monitoring channel and the unit SCADA data are continuously collected at the same time, with a collection interval S and a collection time length U. The time domain audio data of each monitoring channel and the unit SCADA data are taken as a group of data to complete the transmission and storage of the collected data. The unit SCADA data includes the ambient temperature T, yaw angle φ, pitch angle Wind wheel speed ω.

[0028] Preferably, in step 3, the analysis of noise data during the operation of the wind turbine generator set includes the following steps:

[0029] S1. Blade health monitoring data screening: call a set of data in sequence according to the collection time sequence to analyze the pitch angle Time domain data: If the unit starts or stops within the acquisition time length U, the data set will be abandoned; if the unit does not start or stop within the acquisition time length U, the data set will be used as effective data for blade health monitoring, and will be numbered in sequence and recorded as the i-th data set;

[0030] S2. Extraction of effective data parameters for blade health monitoring: Call the data of group i, perform time domain analysis on the SCADA data of the unit within the collection time length U, and obtain the average ambient temperature Average yaw angle Average wind wheel speed Perform spectrum analysis on the time domain audio data of each monitoring channel within the acquisition time length U to obtain the acquisition noise spectrum data results of each monitoring channel {(θ n ,Ar n (f)| i ), n = 1, 2, ..., N}, where Ar n (f) is the frequency domain energy amplitude of the collected noise of monitoring channel #n, f is the frequency, and the subscript i represents the i-th group of data of the unit operation status;

[0031] The frequency domain energy amplitude of the collected noise of monitoring channel #n is Ar n (f)| i Subtract the corresponding ambient noise frequency domain energy amplitude Ac n (f)| stop , the frequency domain energy amplitude expression of the unit noise of monitoring channel #n is defined as:

[0032] A n (f)| i =Ar n (f)| i -Ac n (f)| stop

[0033] Thus, the unit noise spectrum data results of each monitoring channel are obtained {(θ n ,Aw n (f)| i ), n=1, 2,...,N}, where Aw n (f) is the frequency domain energy amplitude of the unit noise of monitoring channel #n, f is the frequency, and the subscript i represents the i-th group of data of the unit operation status;

[0034] The relative angle direction expression of the i-th group of data in the unit operation status is defined as:

[0035]

[0036] in, Analyze the unit noise spectrum data results by the angle interpolation method of adjacent monitoring channels and define any relative angle direction The result function of the unit noise spectrum data Among them, Aw is the frequency domain energy amplitude of the unit noise, f is the frequency, is the relative angle direction, θ is the absolute angle direction, and the subscript i represents the i-th group of data of the unit operation status;

[0037] S3. Calculation of blade noise characteristic distribution state parameters: The result function of the unit noise spectrum data under the condition is defined as the result function of the blade noise spectrum data Wherein, m is the number of wind rotor blades;

[0038] The relative angle direction The value range is [0,2π), which is divided into P angle intervals. In the above equation, take the average value of the relative angle direction The maximum value of energy amplitude in the sum frequency domain The blade noise spectrum data results corresponding to the angle interval are formed, and the blade noise spectrum data results of each angle interval are obtained. Among them, Ab is the frequency domain energy amplitude of blade noise, the subscript i represents the i-th group of data in the unit operation state, and the subscript j represents the j-th angle interval; the parameters of the blade noise characteristic distribution state of the i-th group of data in the unit operation process are as follows, including the average ambient temperature Average wind wheel speed Blade noise spectrum data results Recorded as

[0039] S4. Accumulation of blade noise characteristic distribution state parameters: Repeat S1-S3, traverse each group of unit operation status data, and obtain a series of blade noise characteristic distribution state parameter results during the unit operation process. Among them, i=1, 2,…

[0040] Preferably, in step 4, the evaluation of noise data during the operation of the wind turbine generator set specifically includes the following steps:

[0041] T1. Blade noise characteristic distribution state parameter clustering: Taking a fixed continuous time period as the time range division standard, the accumulated blade noise characteristic distribution state parameter results during the operation of a series of units are divided into several clusters according to the time range they belong to; the zth blade noise characteristic distribution state parameter data cluster is recorded as:

[0042]

[0043] T2, blade noise characteristic distribution state parameter correction: the ambient temperature T range [T B ,T T ), divided into Q temperature intervals, in the kth temperature interval The average ambient temperature is Where k = 1, 2, ..., Q; the wind wheel speed ω range (0, ω T ], divided into R speed intervals, in the rth speed interval The average wind wheel speed is Where r = 1, 2, ..., R;

[0044] For each blade noise frequency domain energy amplitude Ab in the zth blade noise characteristic distribution state parameter data family ij , according to its corresponding average ambient temperature Select the temperature range There is an average ambient temperature T in the temperature range k Corresponding to it; according to its corresponding average wind wheel speed Select the speed range you want to enter There is an average wind wheel speed ω in the speed range r Corresponding to it; it itself has a relative angle direction Correspondingly, all the results in the z-th blade noise characteristic distribution state parameter data family are traversed, so that the composition and the ambient temperature T k 、Wind wheel speedω r and relative angular direction The corresponding blade noise frequency domain energy amplitude data set {Ab ij}| krj，z , where k = 1, 2, ..., Q, r = 1, 2, ..., R, j = 1, 2, ..., P; there are Q·R·P sets of the same blade noise frequency domain energy amplitude data in the z-th blade noise characteristic distribution state parameter data family;

[0045] For each blade noise frequency domain energy amplitude data set {Ab ij}| krj，z The data contained in it conforms to the normal distribution law. Its probability density function is used to eliminate potential measurement error data and analyze its average value The results of the distribution state parameters of the blade noise characteristics after the composition correction Wherein, k = 1, 2, ..., Q, r = 1, 2, ..., R, j = 1, 2, ..., P;

[0046] T3. Construction of blade noise characteristic distribution state function: The corrected blade noise characteristic distribution state parameter result The blade noise characteristic distribution state function expression of the zth blade noise characteristic distribution state parameter data family is defined by the piecewise linear fitting method. Satisfy the independent variable ambient temperature T value range [T B ,T T ), wind wheel speed ω value range (0, ω T], relative angle direction Range [0,2π), all corrected blade noise characteristic distribution state parameter results The corresponding blade noise characteristic distribution state function expression is On the space surface composed of;

[0047] Take the corrected blade noise characteristic distribution state parameter result of the first blade noise characteristic distribution state parameter data group As the initial value of the blade noise characteristic distribution state function Where k = 1, 2, ..., Q, r = 1, 2, ..., R, j = 1, 2, ..., P; for the independent variable coordinate position If exists And excluding the deviation of blade health monitoring device and the cause of blade damage, the value is assigned Locally update the initial value of the blade noise characteristic distribution state function Similarly, using the initial value of the blade noise characteristic distribution state function Define its initial value expression

[0048] T4. Obtain the change in the blade noise characteristic distribution state result: define the step function as:

[0049]

[0050] The change in the blade noise characteristic distribution state result of the zth blade noise characteristic distribution state parameter data family is defined as:

[0051]

[0052] That is, in the non-zero spatial intersection area of ​​the value range of the blade noise characteristic distribution state function expression and the blade noise characteristic distribution state function initial value expression of the z-th blade noise characteristic distribution state parameter data family, the generalized volume weighted average value of the relative error absolute value of the blade noise frequency domain energy amplitude average value is solved, where the three independent variables include ambient temperature T, wind wheel speed ω, relative angle direction The corresponding value ranges are [T B ,T T ),(0,ω T ], [0,2π);

[0053] T5. Blade health status judgment: If the blade noise characteristic distribution state result change amount Δ z is less than the blade health status monitoring technical requirement threshold TD, then the blade is in healthy operation; if the blade noise characteristic distribution state result change amount Δz If it is greater than or equal to the technical requirement threshold TD for blade health status monitoring, it is necessary to arrange on-site inspection, including inspection of the deviation of the blade health monitoring device and inspection of the blade damage, eliminate the influence of environmental factors, find the physical reason why the corresponding change exceeds the technical requirement threshold, and then carry out maintenance work on the blade health monitoring device or the blade itself; after the maintenance work is completed, the collection and analysis of the environmental noise time domain audio data in step R1 shall be carried out again.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] The present invention proposes a wind turbine blade health monitoring device, in which a radial bracket is arranged on the outer wall surface of the tower, and a circumferential bracket is installed in the circumference of the radial bracket, which plays a role in constraining the freedom of the radial bracket and ensuring the stability of the spatial arrangement posture of the monitoring device; the installation position of the radial bracket is higher than the tower door, which can reduce the influence of the tower door on the regular spatial distribution characteristics of noise after coupling with the tower wall; a sound level meter and a protective cover are placed on the radial bracket, and the cantilever length ensures that the placement position of the sound level meter and the protective cover is higher than the installation position of the radial bracket on the tower, which can reduce the influence of the bracket on the spatial distribution characteristics of noise; the installation of the sound level meter and the protective cover can collect audio data and protect the sensor, and prevent the influence or damage of special weather conditions such as wind, frost, rain and snow on the sensor; a plurality of radial brackets, sound level meters and protective covers are arranged and evenly distributed in the circumference of the tower, and the number of radial brackets, sound level meters and protective covers is the same, so as to increase the number of information monitoring channels in different directions and solve the influence of the yaw angle of the unit on the collection data analysis process. The wind turbine blade health monitoring device proposed in the present invention stabilizes the spatial arrangement of the sound level meter and the protective cover through the cooperation of the radial support and the circumferential support, and accurately measures the audio information in multiple directions through the sound level meter and the protective cover. The monitoring device has a simple structure and is easy to install, debug and monitor.

[0056] Furthermore, the radial bracket is connected to the circumferential bracket through the attitude adjustment device, and the radial bracket is connected to the sound level meter and the protective cover through the attitude adjustment device, which plays a role in adjusting the relative spatial position and ensuring the spatial arrangement attitude accuracy of the monitoring device.

[0057] The present invention proposes a monitoring method for a blade health monitoring device of a horizontal axis wind turbine generator set. The method utilizes the characteristics and laws of the blade noise of the wind turbine generator set to analyze the noise characteristic parameters in the frequency domain space, remove the interference of the environmental noise on the unit noise in the collected data, and is conducive to improving the evaluation accuracy of the relative change of noise; extracts the blade noise characteristic parameters from the unit noise, and uses the yaw angle of the wind turbine generator set to locate the spatial distribution law of the blade noise, so as to achieve the fixation of the monitoring device, which is conducive to maintaining the spatial positioning accuracy of the sensor and solving the problem of the effectiveness of the long-term monitoring method of blade noise; introduces three discrete parameters, namely, ambient temperature, wind wheel speed, and relative angle direction, to classify the discrete parameter conditions and evaluate the blade noise characteristics, which is conducive to improving the evaluation accuracy of the change in the blade noise characteristic distribution law. The monitoring method for a blade health monitoring device of a wind turbine generator set proposed by the present invention realizes the health judgment of the blade operation status by evaluating the changes in the blade noise characteristic distribution state parameters during the operation of the unit, and provides technical support for the blade health monitoring work of the wind turbine generator set.

[0058] Furthermore, the normal distribution law of the measurement results is used to eliminate the measurement error data and correct the parameter results of the blade noise characteristic distribution state. The environmental noise collection and update link is designed to effectively remove the interference of environmental noise on the unit noise in the collected data, which is conducive to improving the accuracy of the system's evaluation of the relative change of blade noise.

[0059] Furthermore, the change in the result of the blade noise characteristic distribution state is defined, and the relative change of the blade noise under different parameters is comprehensively counted, so as to improve the resolution of distinguishing the change of the blade's inherent characteristics through noise monitoring and ensure the sensitivity of the blade operation health judgment. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0061] Figure 1 It is a cross-sectional view of the blade health monitoring device of the present invention.

[0062] Figure 2 It is a top view of the blade health monitoring device of the present invention.

[0063] Figure 3 The figure is a flow chart of the blade health monitoring method of the present invention.

[0064] Among them: 1- radial support; 2- circumferential support; 3- attitude adjustment device; 4- sound level meter and protective cover; 5- tower. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0066] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0067] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0068] In the description of the embodiments of the present invention, it should be noted that if the terms "upper", "lower", "horizontal", "inner", etc. indicate an orientation or positional relationship based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the product of the invention is usually placed when in use, it is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In addition, the terms "first", "second", etc. are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0069] In addition, if the term "horizontal" appears, it does not mean that the component must be absolutely horizontal, but can be slightly tilted. For example, "horizontal" only means that its direction is more horizontal than "vertical", which does not mean that the structure must be completely horizontal, but can be slightly tilted.

[0070] In the description of the embodiments of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "set", "install", "connect", and "connect" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal connection of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0071] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0072] The present invention proposes a wind turbine blade health monitoring device, such as Figure 1 and Figure 2 As shown, it includes a radial bracket 1, a circumferential bracket 2, a sound level meter and a protective cover 4, and each part is connected by mechanical assembly; a radial bracket 1 is radially installed on the outer wall of the tower 5, a sound level meter and a protective cover 4 are installed at a relatively high end of the radial bracket 1, and a circumferential bracket 2 is circumferentially installed on the radial bracket 1; a posture adjustment device 3 is provided on the radial bracket 1, the radial bracket 1 is connected to the circumferential bracket 2 through the posture adjustment device 3, the radial bracket 1 is connected to the sound level meter and the protective cover 4 through the posture adjustment device 3, there are several radial brackets 1, sound level meters and protective covers 4, and the number is equal, and the radial brackets 1, the sound level meters and the protective covers 4 are evenly arranged around the tower 5.

[0073] The present invention proposes a wind turbine blade health monitoring device, and the structure and function of each part are described as follows:

[0074] The radial bracket 1 has a structure in the form of a cantilever support structure installed radially along the tower 5. A plurality of radial brackets 1 are evenly arranged around the bottom tower 5 to form a spatial arrangement rotationally symmetrical with the axis of the tower 5. The installation position is higher than the tower door. The radial bracket 1 does not interfere with the blades and is connected to the tower 5 by mechanical assembly. The function of the radial bracket 1 is to place the sound level meter and the protective cover 4. The cantilever length of the radial bracket 1 ensures that the placement position of the sound level meter and the protective cover 4 is higher than the installation position of the radial bracket 1 on the tower 5.

[0075] Circumferential bracket 2, the structural form of circumferential bracket 2 is: an annular support structure, arranged rotationally symmetrically with the axis of tower 5, and mechanically assembled and connected to the radial bracket 1 arranged in space through attitude adjustment device 3; the number of circumferential bracket 2 is one or several; the function of circumferential bracket 2 is: to constrain the circumferential freedom of radial bracket 1 arranged in space.

[0076] The posture adjustment device 3 has a structural form of: connecting fasteners or bolt knobs to form a mechanical assembly connection between the radial bracket 1 and the circumferential bracket 2, the radial bracket 1 and the sound level meter and the protective cover 4; the function of the posture adjustment device 3 is to adjust the relative position between the mechanical assembly connecting parts to ensure the spatial posture accuracy of the radial bracket 1, the circumferential bracket 2, the sound level meter and the protective cover 4.

[0077] The sound level meter and protective cover 4 have the following structural forms: audio acquisition equipment; the number of the sound level meter and protective cover 4 is the same as that of the radial bracket 1; they are mechanically assembled and connected to the radial bracket 1 through the attitude adjustment device 3; multiple sound level meters and protective covers 4 are spatially arranged in a plane perpendicular to the axis of the tower 5, and are evenly distributed in a circumferential position with the intersection of the axis of the tower 5 and the plane as the center; the functions of the sound level meter and protective cover 4 are: to collect audio data of the unit operation, to protect the sensor, and to prevent the influence or damage of special weather conditions such as wind, frost, rain and snow on the sensor.

[0078] The present invention proposes a monitoring method for a blade health monitoring device of a horizontal axis wind turbine generator set. The flow chart of the monitoring method is as follows: Figure 3 As shown:

[0079] Step 1: Installation of wind turbine blade health monitoring device:

[0080] One of the sound level meters and protective covers 4 is located in the wake direction of the unit in the main wind direction during installation, which is defined as monitoring channel #1; the spatial arrangement of several sound level meters and protective covers 4 is located in a plane perpendicular to the axis of the tower 5, and several sound level meters and protective covers 4 are evenly distributed in the circumferential position with the intersection of the axis of the tower 5 and the plane where the several sound level meters and protective covers 4 are located as the center of the circle; a total of N monitoring channels are used, which are arranged in the positive direction of the yaw angle and are defined as monitoring channels #2, #3, ..., #N in sequence; the normal yaw angle of the unit to the wind in the main wind direction during installation is recorded as φ0, and the absolute angle direction θ corresponding to the definition of monitoring channel #n n =φ0+2π(n-1) / N, where n=1, 2, …, N.

[0081] Step 2: Collection of noise data during the operation of wind turbines:

[0082] It includes the debugging of wind turbine blade health monitoring device and the collection of wind turbine operation audio data, realizing the wind turbine blade health monitoring data collection, transmission, storage and call functions.

[0083] The collection of noise data during the operation of wind turbine generator sets specifically includes the following steps:

[0084] R1, blade health monitoring device debugging: check the normal operation of each monitoring channel instrument in turn, and ensure that the measurement data result error is within the technical requirements of quality inspection through calibration; when the unit is in shutdown state and there is no abnormal sound in the environment, continuously collect the time domain audio data of each monitoring channel at the same time, the collection time length U, and complete the transmission and storage of the collected data; call the data for spectrum analysis to obtain the environmental noise spectrum data results of each monitoring channel {(θ n ,Ac n (f)|stop ), n=1, 2, ..., N}, where Ac n (f) is the frequency domain energy amplitude of the ambient noise of monitoring channel #n, f is the frequency, and the subscript stop indicates the shutdown status of the unit.

[0085] R2, wind turbine operation audio data collection: When the unit is in operation, the time domain audio data of each monitoring channel and the unit SCADA data are continuously collected at the same time, with a collection interval period S and a collection time length U. The time domain audio data of each monitoring channel and the unit SCADA data are taken as a group of data to complete the transmission and storage of the collected data. Among them, the unit SCADA data includes the ambient temperature T, yaw angle φ, pitch angle Wind wheel speed ω.

[0086] Step 3: Analysis of noise data during wind turbine operation:

[0087] It includes blade health monitoring data screening, effective data parameter extraction, blade noise characteristic distribution state parameter calculation and accumulation, realizing the wind turbine blade health monitoring data screening and analysis functions.

[0088] The analysis of noise data during the operation of wind turbines specifically includes the following steps:

[0089] S1, blade health monitoring data screening: according to the collection time sequence, call a set of data in turn to analyze the pitch angle For time domain data, if the unit starts or stops within the acquisition time length U, this group of data will be abandoned; if the unit does not start or stop within the acquisition time length U, this group of data will be used as valid data for blade health monitoring, and will be numbered sequentially and recorded as the i-th group of data.

[0090] S2, Extraction of effective data parameters for blade health monitoring: Call the data of group i, perform time domain analysis on the SCADA data of the unit within the collection time length U, and obtain the average ambient temperature Average yaw angle Average wind wheel speed Perform spectrum analysis on the time domain audio data of each monitoring channel within the acquisition time length U to obtain the acquisition noise spectrum data results of each monitoring channel {(θ n ,Ar n (f)| i ), n = 1, 2, ..., N}, where Ar n (f) is the energy amplitude of the collected noise in the frequency domain of monitoring channel #n, f is the frequency, and the subscript i represents the i-th group of data on the unit’s operating status.

[0091] The frequency domain energy amplitude of the collected noise of monitoring channel #n is Arn (f)| i Subtract the corresponding ambient noise frequency domain energy amplitude Ac n (f)| stop , the frequency domain energy amplitude expression of the unit noise of monitoring channel #n is defined as:

[0092] A n (f)| i =Ar n (f)| i -Ac n (f)| stop

[0093] Thus, the unit noise spectrum data results of each monitoring channel are obtained {(θ n ,Aw n (f)| i ), n=1, 2,...,N}, where Aw n (f) is the frequency domain energy amplitude of the unit noise of monitoring channel #n, f is the frequency, and the subscript i represents the i-th group of data of the unit operation status.

[0094] The relative angle direction expression of the i-th group of data in the unit operation status is defined as:

[0095]

[0096] in, Analyze the unit noise spectrum data results by the angle interpolation method of adjacent monitoring channels and define any relative angle direction The result function of the unit noise spectrum data Where Aw is the frequency domain energy amplitude of the unit noise, f is the frequency, is the relative angle direction, θ is the absolute angle direction, and the subscript i represents the i-th group of data on the unit's operating status.

[0097] S3, blade noise characteristic distribution state parameter calculation: frequency The result function of the unit noise spectrum data under the condition is defined as the result function of the blade noise spectrum data Where m is the number of wind rotor blades.

[0098] The relative angle direction The value range is [0,2π), which is divided into P angle intervals. In the above equation, take the average value of the relative angle direction The maximum value of energy amplitude in the sum frequency domain The blade noise spectrum data results corresponding to the angle interval are formed, thereby obtaining the blade noise spectrum data results of each angle interval Among them, Ab is the frequency domain energy amplitude of blade noise, the subscript i represents the i-th group of data of the unit operation status, and the subscript j represents the j-th angle interval.

[0099] The results of the blade noise characteristic distribution state parameters of the i-th group of data during the operation of the unit, including the average ambient temperature Average wind wheel speed Blade noise spectrum data results Record as

[0100] S4, accumulation of blade noise characteristic distribution state parameters: repeat the first three steps, traverse each group of unit operation status data, and obtain a series of blade noise characteristic distribution state parameter results during the unit operation process. Among them, i=1, 2,…

[0101] Step 4: Evaluation of noise data during wind turbine operation:

[0102] It includes the classification and correction of blade noise characteristic distribution state parameters, the construction of blade noise characteristic distribution state function, the acquisition of result change and the judgment of blade operation health status, so as to realize the wind turbine blade health monitoring data evaluation function.

[0103] The evaluation of noise data during the operation of wind turbines specifically includes the following steps:

[0104] T1, blade noise characteristic distribution state parameter clustering: Taking a fixed continuous time period as the time range division standard, the accumulated blade noise characteristic distribution state parameter results during the operation of a series of units are divided into several clusters according to the time range they belong to, and the zth blade noise characteristic distribution state parameter data cluster is recorded as:

[0105]

[0106] T2, blade noise characteristic distribution state parameter correction: the ambient temperature T range [T B ,T T ), divided into Q temperature intervals, in the kth temperature interval The average ambient temperature is Where k = 1, 2, ..., Q; the wind wheel speed ω range (0, ω T ], divided into R speed intervals, in the rth speed interval The average wind wheel speed is Among them, r = 1, 2, …, R.

[0107] For each blade noise frequency domain energy amplitude Ab in the zth blade noise characteristic distribution state parameter data family ij, according to its corresponding average ambient temperature Select the temperature range There is an average ambient temperature T in the temperature range k Corresponding to it; according to its corresponding average wind wheel speed Select the speed range you want to enter There is an average wind wheel speed ω in the speed range r Corresponding to it; it itself has a relative angle direction Correspondingly, all the results in the z-th blade noise characteristic distribution state parameter data family are traversed, so that the composition and the ambient temperature T k 、Wind wheel speedω r and relative angular direction The corresponding blade noise frequency domain energy amplitude data set {Ab ij}| krj，z , where k = 1, 2, ..., Q, r = 1, 2, ..., R, j = 1, 2, ..., P; there are Q·R·P sets of the same blade noise frequency domain energy amplitude data in the z-th blade noise characteristic distribution state parameter data family.

[0108] For each blade noise frequency domain energy amplitude data set {Ab ij}| krj，z The data contained in it conforms to the normal distribution law. Its probability density function is used to eliminate potential measurement error data and analyze its average value The results of the distribution state parameters of the blade noise characteristics after the composition correction Among them, k=1, 2, ..., Q, r=1, 2, ..., R, j=1, 2, ..., P.

[0109] T3, blade noise characteristic distribution state function construction: the corrected blade noise characteristic distribution state parameter result The blade noise characteristic distribution state function expression of the zth blade noise characteristic distribution state parameter data family is defined by the piecewise linear fitting method. Satisfy the independent variable ambient temperature T value range [T B ,T T ), wind wheel speed ω value range (0, ω T ], relative angle direction Range [0,2π), all corrected blade noise characteristic distribution state parameter results The corresponding blade noise characteristic distribution state function expression is On the composed space surface.

[0110] Take the corrected blade noise characteristic distribution state parameter result of the first blade noise characteristic distribution state parameter data group As the initial value of the blade noise characteristic distribution state function Where k = 1, 2, ..., Q, r = 1, 2, ..., R, j = 1, 2, ..., P; for the independent variable coordinate position If exists And excluding the deviation of blade health monitoring device and the cause of blade damage, the value is assigned Locally update the initial value of the blade noise characteristic distribution state function Similarly, using the initial value of the blade noise characteristic distribution state function Define its initial value expression

[0111] T4, obtain the change in the blade noise characteristic distribution state result: define the step function as:

[0112]

[0113] The change in the blade noise characteristic distribution state result of the zth blade noise characteristic distribution state parameter data family is defined as:

[0114]

[0115] That is, in the non-zero spatial intersection area of ​​the value range of the blade noise characteristic distribution state function expression and the blade noise characteristic distribution state function initial value expression of the z-th blade noise characteristic distribution state parameter data family, the generalized volume weighted average value of the relative error absolute value of the blade noise frequency domain energy amplitude average value is solved, where the three independent variables include ambient temperature T, wind wheel speed ω, relative angle direction The corresponding value ranges are [T B ,T T ),(0,ω T ]、[0,2π).

[0116] T5, blade health status judgment: if the blade noise characteristic distribution state result change amount Δ z is less than the blade health status monitoring technical requirement threshold TD, then the blade is in healthy operation; if the blade noise characteristic distribution state result change amount Δ z If it is greater than or equal to the technical requirement threshold TD for blade health status monitoring, it is necessary to arrange on-site inspection, including inspection of blade health monitoring device deviation and blade damage, eliminate the influence of environmental factors, find the physical reason why the corresponding change exceeds the technical requirement threshold, and then carry out maintenance work on the blade health monitoring device or the blade itself; after the maintenance work is completed, it is necessary to re-carry out the collection and analysis of environmental noise time domain audio data in step R1 of step 2.

[0117] This method effectively realizes the health monitoring of wind turbine blades, including design, testing, and analysis, forming a complete set of wind turbine blade health monitoring devices and monitoring methods.

[0118] The present invention proposes a wind turbine blade health monitoring device and monitoring method. The device supports sensor devices through the cooperation of radial supports and circumferential supports, adjusts the spatial posture of the sensor devices through an attitude adjustment device, and simultaneously measures audio information in multiple directions and protects the sensor devices through a sound level meter and a protective cover, thereby realizing the wind turbine blade health monitoring data acquisition function. The method analyzes noise characteristic parameters in the frequency domain space, removes the interference of environmental noise on the unit noise from the collected data, extracts blade noise information using the wind rotor speed, and locates the spatial distribution law of blade noise characteristic parameters using the yaw angle when the wind turbine is running. The method uses discrete ambient temperature, wind rotor speed, and phase For the angular direction parameters, the frequency domain energy amplitude under the discrete parameter combination is extracted to form the blade noise characteristic distribution state parameter result; the normal distribution law of the data result under the discrete parameter combination is used to eliminate the influence of the measurement error, correct the blade noise characteristic distribution state parameter result, and realize the wind turbine blade health monitoring data analysis function; the discrete parameter value range integral is used to define the change amount of the blade noise characteristic distribution state result, and realize the wind turbine blade health monitoring data evaluation function; the device and method realize the monitoring and judgment of the blade operation health status through the collection, analysis and evaluation of noise-related data during the operation of the wind turbine, and provide technical support for the wind turbine blade health monitoring work.

[0119] The present invention proposes a wind turbine blade health monitoring device and monitoring method, which plays a role in collecting and evaluating the noise changes caused by blade damage or failure, and lays a foundation for remote monitoring of the healthy operation status of wind turbines and formulating operation and maintenance work cycles and plans. The design and development process of this method is clear and definite, the implementation link is easy to monitor and measure, and the data analysis link is easy to statistically process, which can effectively reduce the cycle and cost of blade health monitoring work. It has the following advantages: 1) It effectively combines device design, experimental testing, and data analysis, and uses the high efficiency of the platform for data integration to improve the timeliness and reliability of the analysis and evaluation process, and coordinates and guarantees from design, testing, analysis, and application to ultimately achieve refined evaluation and judgment of blade health, providing technical support for wind turbine blade health monitoring, and forming a complete set of wind turbine blade health monitoring devices and monitoring methods; 2) It defines the change in the result of the blade noise characteristic distribution state, comprehensively statistics the relative change of blade noise under different parameters, improves the resolution of identifying changes in the inherent characteristics of the blade itself through noise monitoring, and ensures the sensitivity of blade health judgment; 3) It uses the yaw angle during the operation of the wind turbine to locate the spatial distribution law of blade noise, which can realize the fixation of the blade health monitoring device, which is conducive to improving the spatial positioning accuracy of the sensor, and is conducive to improving the evaluation accuracy of the change in the spatial distribution law of blade noise after the yaw angle changes during the normal operation of the unit; 4) It introduces three parameters: ambient temperature, wind wheel speed, and relative angle direction, to analyze the data 5) Effectively classify the blade noise according to different parameter conditions to evaluate the blade noise characteristics in detail, which is conducive to improving the evaluation accuracy of the change in the blade noise characteristic distribution law; 6) Effectively transform the collected time domain signal into the frequency domain space for analysis according to the characteristics and laws of the wind turbine blade noise, which is conducive to obtaining the noise characteristic parameters, especially extracting the blade noise characteristic parameters from the unit noise; 7) Design the environmental noise collection and update link to effectively remove the interference of environmental noise on the unit noise in the collected data, which is conducive to improving the accuracy of the relative change of blade noise in the later system evaluation; 8) Design the spatial arrangement of the blade health monitoring device, taking into account the regularity of the noise after coupling with the tower wall, effectively record the spatial distribution of blade noise, increase the amount of collected evaluation data, and improve the description dimension of the results; effectively use the normal distribution law of the measurement results, eliminate the measurement error data, and correct the parameter results of the blade noise characteristic distribution state, which is conducive to improving the evaluation accuracy of the change in the blade noise characteristic distribution law; 9) Use the yaw angle when the wind turbine is running to locate the blade noise spatial distribution law, realize the fixation of the blade health monitoring device, and improve the spatial positioning accuracy of the sensor.

[0120] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A monitoring method for a wind turbine blade health monitoring device, characterized in that: A wind turbine blade health monitoring device is used, comprising a radial support (1), a circumferential support (2), and a sound level meter and a protective cover (4); the radial support (1) is a cantilever support structure, one end of the radial support (1) is radially mounted on the outer side wall of a tower (5) and is evenly arranged around the tower (5), and the installation position of the radial support (1) is higher than the tower door; the sound level meter and the protective cover (4) are mounted on the other end of the radial support (1); there are a plurality of radial supports (1), a plurality of sound level meters and a plurality of protective covers (4); the number of radial supports (1), a plurality of sound level meters and a plurality of protective covers (4) is equal to or greater than 100%. equal; the circumferential support (2) is an annular support structure, the circumferential support (2) is mounted on the radial support (1) along the circumferential direction; the radial support (1) is also provided with a posture adjustment device (3), the radial support (1) and the circumferential support (2) are connected via the posture adjustment device (3); the radial support (1) is connected to the sound level meter and the protective cover (4) via the posture adjustment device (3); the posture adjustment device (3) is a connecting fastener or a bolt knob; the circumferential support (2) is arranged rotationally symmetrically with respect to the axis of the tower (5); the number of the circumferential support (2) is several; The monitoring method comprises the following steps: Step 1: Installation of wind turbine blade health monitoring device: One of the sound level meters and protective covers (4) is located in the wake direction of the unit in the main wind direction during installation, and is defined as monitoring channel #1; the spatial arrangement of the plurality of sound level meters and protective covers (4) is located in a plane perpendicular to the axis of the tower (5), and the plurality of sound level meters and protective covers (4) are evenly distributed in the circumferential position with the intersection of the axis of the tower (5) and the plane where the plurality of sound level meters and protective covers (4) are located as the center of the circle; a total of N monitoring channels are used, which are arranged in the positive direction of the yaw angle and are defined as monitoring channels #2, #3, ..., #N in sequence; the normal yaw angle of the unit to the wind in the main wind direction during installation is recorded as φ0, and the absolute angle direction θ corresponding to the monitoring channel #n is defined n =φ0+2π(n-1) / N, where n=1, 2, ..., N; Step 2: Collection of noise data during the operation of wind turbines: It includes the debugging of the wind turbine blade health monitoring device and the collection of wind turbine operation audio data, realizing the functions of wind turbine blade health monitoring data collection, transmission, storage and call; Step 3: Analysis of noise data during wind turbine operation: It includes blade health monitoring data screening, effective data parameter extraction, blade noise characteristic distribution state parameter calculation and accumulation, realizing the wind turbine blade health monitoring data screening and analysis functions; Step 4: Evaluation of noise data during wind turbine operation: It includes the classification and correction of blade noise characteristic distribution state parameters, the construction of blade noise characteristic distribution state function, the acquisition of result change and the judgment of blade operation health state, so as to realize the function of wind turbine blade health monitoring data evaluation; In step 3, the analysis of noise data during the operation of the wind turbine generator set includes the following steps: S1. Blade health monitoring data screening: call a set of data in sequence according to the collection time sequence to analyze the pitch angle Time domain data: If the unit starts or stops within the acquisition time length U, the data set will be abandoned; if the unit does not start or stop within the acquisition time length U, the data set will be used as effective data for blade health monitoring, and will be numbered in sequence and recorded as the i-th data set; S2. Extraction of effective data parameters for blade health monitoring: Call the data of group i, perform time domain analysis on the SCADA data of the unit within the collection time length U, and obtain the average ambient temperature Average yaw angle Average wind wheel speed Perform spectrum analysis on the time domain audio data of each monitoring channel within the acquisition time length U to obtain the acquisition noise spectrum data results of each monitoring channel {(θ n ,Ar n (f) i ), n = 1, 2, ..., N}, where Ar n (f) is the frequency domain energy amplitude of the collected noise of monitoring channel #n, f is the frequency, and the subscript i represents the i-th group of data of the unit operation status; The frequency domain energy amplitude of the collected noise of monitoring channel #n is Ar n (f) i Subtract the corresponding ambient noise frequency domain energy amplitude Ac n (f) stop , the frequency domain energy amplitude expression of the unit noise of monitoring channel #n is defined as: <h2 style=";text-align:left;direction:ltr">Aw<h2 style=";text-align:left;direction:ltr"> n <h2 style=";text-align:left;direction:ltr"> (f)<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> =Ar<h2 style=";text-align:left;direction:ltr"> n <h2 style=";text-align:left;direction:ltr"> (f)<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> -Ac<h2 style=";text-align:left;direction:ltr"> n <h2 style=";text-align:left;direction:ltr"> (f)<h2 style=";text-align:left;direction:ltr"> stop Thus, the unit noise spectrum data results of each monitoring channel are obtained {(θ n ,Aw n (f) i ), n=1, 2,...,N}, where Aw n (f) is the frequency domain energy amplitude of the unit noise of monitoring channel #n, f is the frequency, and the subscript i represents the i-th group of data of the unit operation status; The relative angle direction expression of the i-th group of data in the unit operation status is defined as: in, Analyze the unit noise spectrum data results by the angle interpolation method of adjacent monitoring channels and define any relative angle direction The result function of the unit noise spectrum data Where Aw is the frequency domain energy amplitude of the unit noise, f is the frequency, is the relative angle direction, θ is the absolute angle direction, and the subscript i represents the i-th group of data of the unit operation status; S3. Calculation of blade noise characteristic distribution state parameters: The result function of the unit noise spectrum data under the condition is defined as the result function of the blade noise spectrum data Wherein, m is the number of wind rotor blades; The relative angle direction The value range is [0,2π), which is divided into P angle intervals. Take the average value of the relative angle direction The maximum value of energy amplitude in the sum frequency domain The blade noise spectrum data results corresponding to the angle interval are formed, and the blade noise spectrum data results of each angle interval are obtained. Among them, Ab is the frequency domain energy amplitude of blade noise, the subscript i represents the i-th group of data in the unit operation state, and the subscript j represents the j-th angle interval; the parameters of the blade noise characteristic distribution state of the i-th group of data in the unit operation process are as follows, including the average ambient temperature Average wind wheel speed Blade noise spectrum data results Recorded as S4. Accumulation of blade noise characteristic distribution state parameters: Repeat S1-S3, traverse each group of unit operation status data, and obtain a series of blade noise characteristic distribution state parameter results during the unit operation process. Wherein, i=1, 2, ...; In step 4, the evaluation of noise data during the operation of the wind turbine generator set specifically includes the following steps: T1. Blade noise characteristic distribution state parameter clustering: Taking a fixed continuous time period as the time range division standard, the accumulated blade noise characteristic distribution state parameter results during the operation of a series of units are divided into several clusters according to the time range they belong to; the zth blade noise characteristic distribution state parameter data cluster is recorded as τ zs represents the starting time of the zth blade noise characteristic distribution state parameter data cluster, τ ze Indicates the end time of the zth blade noise characteristic distribution state parameter data cluster; T2, blade noise characteristic distribution state parameter correction: the ambient temperature T range [T B ,T T ), divided into Q temperature intervals, in the kth temperature interval The average ambient temperature is Where k = 1, 2, ..., Q; the wind wheel speed ω range (0, ω T ], divided into R speed intervals, in the rth speed interval The average wind wheel speed is Where r = 1, 2, ..., R; For each blade noise frequency domain energy amplitude Ab in the zth blade noise characteristic distribution state parameter data family ij , according to its corresponding average ambient temperature Select the temperature range There is an average ambient temperature T in the temperature range k Corresponding to it; according to its corresponding average wind wheel speed Select the speed range you want to enter There is an average wind wheel speed ω in the speed range r Corresponding to it; it itself has a relative angle direction Correspondingly, all the results in the z-th blade noise characteristic distribution state parameter data family are traversed, so that the composition and the ambient temperature T k 、Wind wheel speedω r and relative angular direction The corresponding blade noise frequency domain energy amplitude data set {Ab ij }| krj，z , where k = 1, 2, ..., Q, r = 1, 2, ..., R, j = 1, 2, ..., P; there are Q·R·P sets of the same blade noise frequency domain energy amplitude data in the z-th blade noise characteristic distribution state parameter data family; For each blade noise frequency domain energy amplitude data set {Ab ij }| krj，z The data contained in it conforms to the normal distribution law. Its probability density function is used to eliminate potential measurement error data and analyze its average value The results of the distribution state parameters of the blade noise characteristics after the composition correction Wherein, k = 1, 2, ..., Q, r = 1, 2, ..., R, j = 1, 2, ..., P; T3. Construction of blade noise characteristic distribution state function: The corrected blade noise characteristic distribution state parameter result The blade noise characteristic distribution state function expression of the zth blade noise characteristic distribution state parameter data family is defined by the piecewise linear fitting method. Satisfy the independent variable ambient temperature T value range [T B ,T T ), wind wheel speed ω value range (0, ω T ], relative angle direction Range [0,2π), all corrected blade noise characteristic distribution state parameter results The corresponding blade noise characteristic distribution state function expression is On the composed space surface; Take the corrected blade noise characteristic distribution state parameter result of the first blade noise characteristic distribution state parameter data group As the initial value of the blade noise characteristic distribution state function Among them, k = 1, 2, ..., Q, r = 1, 2, ..., R, j = 1, 2, ..., P; for the independent variable coordinate position If exists And excluding the deviation of blade health monitoring device and the cause of blade damage, the value is assigned Locally update the initial value of the blade noise characteristic distribution state function Similarly, using the initial value of the blade noise characteristic distribution state function Define its initial value expression T4. Obtain the change in the blade noise characteristic distribution state result: define the step function as: The change in the blade noise characteristic distribution state result of the zth blade noise characteristic distribution state parameter data family is defined as: That is, in the non-zero spatial intersection area of ​​the value range of the blade noise characteristic distribution state function expression and the blade noise characteristic distribution state function initial value expression of the z-th blade noise characteristic distribution state parameter data family, the generalized volume weighted average value of the relative error absolute value of the blade noise frequency domain energy amplitude average value is solved, where the three independent variables include ambient temperature T, wind wheel speed ω, relative angle direction The corresponding value ranges are [T B ,T T ),(0,ω T ], [0,2π); is the average value of the frequency domain energy amplitude of the blade noise of the zth blade noise characteristic distribution state parameter data family, is the initial value of the average value of blade noise frequency domain energy amplitude; T5. Blade health status judgment: If the blade noise characteristic distribution state result change amount Δ z is less than the blade health status monitoring technical requirement threshold TD, then the blade is in healthy operation; if the blade noise characteristic distribution state result change amount Δ z If it is greater than or equal to the technical requirement threshold TD for blade health status monitoring, it is necessary to arrange on-site inspection, including inspection of blade health monitoring device deviation and blade damage, eliminate the influence of environmental factors, find the physical reason why the corresponding change exceeds the technical requirement threshold, and then carry out maintenance work on the blade health monitoring device or the blade itself; after the maintenance work is completed, re-carry out the collection and analysis of environmental noise time domain audio data in steps 2 and 3.

2. The monitoring method of the wind turbine blade health monitoring device according to claim 1, characterized in that: In step 2, the collection of noise data during the operation of the wind turbine generator set includes the following steps: R1. Blade health monitoring device debugging: When the unit is in shutdown state and there is no abnormal sound in the environment, the time domain audio data of each monitoring channel is continuously collected for a collection time length U, and the collection data is transmitted and stored; the data is called for spectrum analysis to obtain the environmental noise spectrum data results of each monitoring channel {(θ n ,Ac n (f) stop ), n=1, 2, ..., N}, where Ac n (f) is the frequency domain energy amplitude of the ambient noise of monitoring channel #n, f is the frequency, and the subscript stop indicates the shutdown state of the unit; R2. Wind turbine operation audio data collection: When the unit is in operation, the time domain audio data of each monitoring channel and the unit SCADA data are continuously collected at the same time, with a collection interval S and a collection time length U. The time domain audio data of each monitoring channel and the unit SCADA data are taken as a group of data to complete the transmission and storage of the collected data. The unit SCADA data includes the ambient temperature T, yaw angle φ, pitch angle Wind wheel speed ω.

Images