Monitoring system of wind driven generator

Abstract

The invention provides a monitoring system of a wind driven generator, which relates to the technical field of wind power generation, and comprises a feature acquisition module used for acquiring videos around the wind driven generator; the data storage and comparison module is used for comparing and matching the data of the feature acquisition module through a trained big data comparison and matching model, extracting a flying bird track in the data after identification and matching, and confirming an impact probability; the monitoring early warning module is used for constructing an impact damage monitoring matrix of the wind driven generator, collecting impact data, judging whether the wind driven generator needs early warning prompt or not according to the impact monitoring matrix, and judging whether to report damage early warning or not based on the characteristic condition of an impact area; and the cooperative recording module is used for transmitting an instruction to adjust the running state of the wind driven generator and storing the monitoring record and the instruction record of the monitoring and early warning module. The invention solves the problem that the wind driven generator lacks a monitoring system for bird collision and cannot timely know and monitor the collision damage of the bird collision to the blades of the wind driven generator.

Description

Technical Field

[0001] The invention belongs to the field of wind power generation, and in particular relates to a monitoring system for a wind power generator. Background Art

[0002] In the early 1970s, due to the "oil crisis", energy shortages emerged. People realized the instability and limitation of conventional mineral energy supply, so seeking clean renewable energy became an important issue in the modern world. Wind energy, as a renewable and pollution-free natural energy, has once again attracted people's attention. People use wind power to drive the windmill blades to rotate, and then increase the speed of rotation through a speed increaser to enable wind turbines to generate electricity.

[0003] However, during the production process, the blades of wind turbines may collide with birds during rotation, causing damage to the blades. Traditional wind turbines lack a bird collision monitoring system and are unable to timely understand and monitor the collision between birds and wind turbines. Summary of the invention

[0004] The present invention provides a monitoring system for a wind turbine generator, which is used to solve at least one technical problem raised in the above background technology.

[0005] In order to solve the above technical problems, the present invention discloses a monitoring system for a wind turbine, comprising: a feature acquisition module, for acquiring a video around the wind turbine;

[0006] The data storage and comparison module is used to store and record the bird posture data in advance, compare and match the feature acquisition module data through the trained big data comparison and matching model, extract the bird trajectory in the feature acquisition module data after identification and matching, and confirm the collision probability;

[0007] The monitoring and early warning module is used to build a wind turbine impact damage monitoring matrix, collect impact data based on the bird impact probability of the data storage comparison module, and determine whether the wind turbine needs an early warning according to the impact monitoring matrix. At the same time, it marks the impact area, continuously monitors the impact area, and determines whether to report a damage warning based on the characteristics of the impact area;

[0008] The collaborative recording module is used to adjust the operating state of the wind turbine according to the instructions transmitted by the monitoring and early warning module, and store the monitoring records and instruction records of the monitoring and early warning module.

[0009] Preferably, the feature data storage and comparison module includes:

[0010] A video frame slicing unit, used for repeatedly slicing the initial video of a preset time length acquired by the video acquisition unit frame by frame;

[0011] The target tracking and comparison unit is used to input a number of slice single frames based on a preset time length into a comparison algorithm model to identify a moving target;

[0012] A screening storage unit is used to pre-store and record the bird posture data, and temporarily store the target slice single frame that exists in the target tracking unit to identify the target, and delete the remaining target slice single frame that does not exist in the tracking unit to identify the target;

[0013] The trajectory analysis unit is used to determine the flight trajectory of the bird based on the comparison and recognition data of the target tracking and comparison unit, and analyze the probability of the bird's flight trajectory hitting the wind turbine blades.

[0014] Preferably, the target tracking and comparison unit includes:

[0015] A pixel processing component is used to obtain the pixel value of each pixel point of a plurality of slice single frames of the initial video of a preset time length collected by the video acquisition unit, take the lower left corner of each slice single frame as the origin, store the pixel values ​​of all the pixel points on each slice single frame into a pixel array according to the coordinate values, and store the pixel arrays corresponding to the plurality of slice single frames into a linked list in chronological order;

[0016] The target tracking component is used to compare the pixel values ​​of the pixel points at the same coordinate position in the pixel arrays of adjacent slice single frames in sequence, store the coordinate values ​​of all the pixel points with pixel values ​​difference between each group of adjacent slice single frames in the position array, and store the position arrays of several groups of adjacent slice single frames in the position linked list in chronological order;

[0017] The target comparison component is used to extract the tracking target data features based on the target tracking component, and compare it with the pre-stored recorded bird posture feature data to determine whether the tracking target is a flying bird;

[0018] The trajectory extraction component is used to accumulate the X-axis coordinate value and the Y-axis coordinate value in each position array of the tracked target in the position linked list and divide them by the number of coordinate values ​​to obtain a number of target overall coordinate values, and store the several target overall coordinate values ​​in the overall position linked list in chronological order.

[0019] Preferably, the monitoring and early warning module includes:

[0020] A wind turbine impact detection and alarm unit is used to obtain the real-time impact situation of the wind turbine within a preset monitoring period based on the wind turbine impact matrix, and determine whether an early warning reminder of the impact is required based on the real-time impact situation;

[0021] The wind turbine blade damage detection and alarm unit is used to evaluate the impact damage of the wind turbine blades within a preset monitoring period, and determine whether to alarm based on the impact damage of the wind turbine blades.

[0022] Preferably, the wind turbine collision detection alarm unit comprises:

[0023] The vibration frequency collection module is used to collect the vibration frequency generated by the collision between birds and blades during the working state of the wind turbine in each preset monitoring period, and classify the frequency level according to the vibration frequency characteristics;

[0024] A vibration frequency monitoring matrix building module is used to build a vibration frequency monitoring matrix A of the entire blade of the wind turbine generator based on the vibration frequency generated when birds collide with the blades during the working state of the wind turbine generator in each preset monitoring period;

[0025] The impact trace collection module is used to collect the traces of birds hitting the blades during the working state of the wind turbine in each preset monitoring period, and classify the trace levels according to the characteristics of the traces;

[0026] The impact trace monitoring matrix building module is used to build a trace monitoring matrix B of the wind turbine blades based on the traces produced by the collision between birds and blades during the working state of the wind turbine in each monitoring cycle;

[0027] A wind turbine impact monitoring matrix construction module is used to construct a wind turbine blade impact monitoring matrix C based on a wind turbine blade vibration frequency monitoring matrix A and a wind turbine blade trace monitoring matrix B within each preset monitoring period;

[0028] The alarm evaluation module is used to evaluate whether an impact alarm is needed based on the evaluation index of the impact monitoring matrix C of the wind turbine blades in each preset monitoring cycle.

[0029] Preferably, based on the vibration frequency generated when birds collide with blades during the working state of the wind turbine in each preset monitoring period, a vibration frequency monitoring matrix A of the wind turbine blades is constructed:

[0030] Among them, A t is the vibration frequency monitoring matrix of the wind turbine blades in the tth preset monitoring period, a 11 is the vibration frequency of the wind turbine blade 11 area, a 12 is the vibration frequency of the wind turbine blade 12 area, a 1i is the vibration frequency of the wind turbine blade 1i area, a 21 is the vibration frequency of the wind turbine blade 21 area, a 22 is the vibration frequency of the wind turbine blade 22 area, a 2i is the vibration frequency of the wind turbine blade 2i area, a 31is the vibration frequency of the wind turbine blade 31 area, a 32 is the vibration frequency of the wind turbine blade 32 area, a 3i is the vibration frequency of the wind turbine blade 3i area, a 41 is the vibration frequency of the wind turbine blade 41 region, a 42 is the vibration frequency of the wind turbine blade 42 region, a 4i is the vibration frequency of the wind turbine blade 4i area, a i1 is the vibration frequency of the wind turbine blade i1 region, a i2 is the vibration frequency of the i2 region of the wind turbine blade, a ii is the vibration frequency of the wind turbine blade region ii;

[0031] Based on the traces of birds colliding with blades during the working state of wind turbines, the impact trace monitoring matrix B of wind turbine blades is constructed:

[0032] Among them, B t is the vibration frequency monitoring matrix of the wind turbine blades in the tth preset monitoring period, b 11 is the impact mark characteristic of the wind turbine blade 11 area, b 12 is the impact mark characteristic of the wind turbine blade 12 area, b 1j b is the impact mark feature of the wind turbine blade 1j area, b 21 is the impact mark characteristic of the wind turbine blade 21 area, b 22 is the impact mark characteristic of the wind turbine blade 22 area, b 2j b is the impact mark feature of the wind turbine blade 2j area, b 31 is the impact mark characteristic of the wind turbine blade 31 area, b 32 is the impact mark characteristic of the wind turbine blade 32 area, b 3j b is the impact mark feature of the 3j area of ​​the wind turbine blade, 41 is the impact mark characteristic of the wind turbine blade 41 area, b 42 is the impact mark characteristic of the wind turbine blade 42 area, b 4j b is the impact mark feature of the 4j area of ​​the wind turbine blade, j1 is the impact mark feature of the wind turbine blade j1 area, b j2 is the impact mark feature of the wind turbine blade j2 area, b jj It is the impact mark characteristic of the jj area of ​​the wind turbine blade.

[0033] Preferably, the impact monitoring matrix C of the wind turbine blades is constructed based on the vibration frequency monitoring matrix A of the wind turbine blades and the impact mark monitoring matrix B of the wind turbine blades:

[0034] Among them, C t is the impact monitoring matrix of the wind turbine blade in the tth preset monitoring period, λ is the matching weight of the vibration frequency of the wind turbine blade, and τ is the matching weight of the impact mark of the wind turbine blade;

[0035] When C t When the value of any element in the collision monitoring matrix exceeds the preset rated value, C t The subscript information and value of the element in the collision monitoring matrix are output and an early warning prompt is given.

[0036] Preferably, the wind turbine blade damage detection and alarm unit comprises:

[0037] The blade damage capture module is used to compare the real-time contour data of the bird impact position of the wind turbine blade with the original contour data based on the bird impact position determined by the wind turbine impact detection alarm unit. If the real-time contour data and the original contour data do not match, it is preliminarily determined that the blade has been damaged by impact and the coordinate data of the mismatched position is captured;

[0038] The blade damage confirmation module is used to extract the pixels and features of the damaged parts of the blade based on the impact damage location coordinates determined by the blade damage capture module, and input them into the damage resolution model trained with various damage feature data to judge and screen the blade damage.

[0039] Preferably, it also includes:

[0040] The blade damage assessment module is used to calculate the area of ​​collision damage and assess the hazard level of collision damage to the blade based on the type judgment and feature data of the collision damage confirmed by the blade damage confirmation module;

[0041] The damage expansion calculation module is used to evaluate the damage expansion coefficient of the current collision damage of the blade when the wind turbine continues to work based on the collision damage area and collision damage hazard level of the blade damage assessment module;

[0042] The alarm level calculation unit is used to obtain the current blade damage alarm level based on the damage expansion coefficient evaluated by the blade performance damage calculation module.

[0043] Preferably, the damage expansion coefficient of the current wind turbine blade is:

[0044] Among them, E dis the damage expansion coefficient of the wind turbine blade at the current moment, ρ is the air density, A is the blade collision damage area, v is the blade rotation speed, K is the action coefficient, ε is the blade collision damage drag coefficient, μ is the mass conversion coefficient, and S is the blade collision damage circumference;

[0045] Current blade damage alarm level:

[0046] Among them, Β is the damage alarm level of the wind turbine blade at the current moment, o is the damage basic coefficient, and Δ is the collision damage hazard level. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0048] Figure 1 This is a schematic diagram of a wind turbine monitoring system proposed by the present invention. DETAILED DESCRIPTION

[0049] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0050] In addition, in the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes, and do not specifically refer to the order or sequence, nor are they used to limit the present invention. They are only used to distinguish components or operations described with the same technical terms, and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions and technical features between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0051] The present invention provides the following embodiments

[0052] Example 1

[0053] The embodiment of the present invention provides a monitoring system for a wind turbine generator, such as Figure 1 As shown, it includes: a feature acquisition module, used to collect videos around the wind turbine;

[0054] The data storage and comparison module is used to store and record the bird posture data in advance, compare and match the feature acquisition module data through the trained big data comparison and matching model, extract the bird trajectory in the feature acquisition module data after identification and matching, and confirm the collision probability;

[0055] The monitoring and early warning module is used to build a wind turbine impact damage monitoring matrix, collect impact data based on the bird impact probability of the data storage comparison module, and determine whether the wind turbine needs an early warning according to the impact monitoring matrix. At the same time, it marks the impact area, continuously monitors the impact area, and determines whether to report a damage warning based on the characteristics of the impact area;

[0056] The collaborative recording module is used to adjust the operating state of the wind turbine according to the instructions transmitted by the monitoring and early warning module, and store the monitoring records and instruction records of the monitoring and early warning module.

[0057] In this embodiment, the area around the wind turbine generator includes a spherical space with the blades of the wind turbine generator as the radius.

[0058] In this embodiment, the trained big data comparison and matching model uses a large amount of flying bird posture data as input model data in advance, and uses the part of the video containing the flying bird as output data to obtain a data model after large model training.

[0059] In this embodiment, a large number of monitoring systems are installed on a single wind turbine. When the wind turbine is working, the monitoring system monitors the real-time video of the spherical space within the radius of the wind turbine blades. The monitoring system is connected to the video integration box via a cable, and the monitoring video is transmitted to the main control integration inside the wind turbine through the video integration box.

[0060] The working principle and beneficial effects of the above technical solution are as follows: the wind turbine monitoring system of the present invention obtains the original video of the spherical space within the radius of the wind turbine blade through the feature acquisition module, and after the data storage and comparison module processes the original video, it identifies the part of the video with the flying bird and stores the video and extracts the flight trajectory of the flying bird, and judges the probability of the flying bird colliding with the wind turbine blade. The monitoring and early warning module monitors the vibration frequency monitoring matrix and the impact trace monitoring matrix of various parts of the wind turbine blade, collects various data of the flying bird at the time of collision, and judges whether the threat of the flying bird colliding with the wind turbine blade needs to be issued an impact alarm and a blade collision damage alarm, and monitors the blade damage at the same time, and judges whether the collision damage of the blade in the working state is further aggravated according to the real-time situation; and issues an alarm when there is aggravated collision damage;

[0061] The monitoring system can monitor the collision between wind turbine blades and birds in real time, and promptly detect the damage caused by the collision between birds and wind turbine blades, thus ensuring the long-term and stable operation of wind turbine blades.

[0062] The present invention can automatically monitor the collision between birds and wind turbine blades, and use data models to analyze in real time the probability of birds hitting wind turbines, the location of birds hitting wind turbines and the damage caused, thereby effectively improving the maintenance efficiency of wind turbine blades and increasing the service life of wind turbine blades.

[0063] Example 2

[0064] Based on Example 1, the feature data storage comparison module includes:

[0065] A video frame slicing unit, used for repeatedly slicing the initial video of a preset time length acquired by the video acquisition unit frame by frame;

[0066] The target tracking and comparison unit is used to input a number of slice single frames based on a preset time length into a comparison algorithm model to identify a moving target;

[0067] A screening storage unit is used to pre-store and record the bird posture data, and temporarily store the target slice single frame that exists in the target tracking unit to identify the target, and delete the remaining target slice single frame that does not exist in the tracking unit to identify the target;

[0068] The trajectory analysis unit is used to determine the flight trajectory of the bird based on the comparison and recognition data of the target tracking and comparison unit, and analyze the probability of the bird's flight trajectory hitting the wind turbine blades.

[0069] In this embodiment, the comparison algorithm model takes several sliced ​​single frames of the original video as input, compares them with adjacent frames in the several sliced ​​single frames, and takes the sliced ​​single frames with differences in the comparison as output.

[0070] The working principle and beneficial effects of the above technical solution are as follows: the video frame slicing unit is used to repeatedly slice the original video of a certain length of time into a number of single frames, and the number of sliced ​​single frames are input into the comparison algorithm model. The comparison algorithm model identifies the moving target and determines whether it is a flying bird. The screening storage unit saves the sliced ​​single frames that pass the bird determination and deletes the sliced ​​single frames that do not pass the bird determination. Then, the moving trajectory of the flying bird is determined based on the sliced ​​single frames that pass the bird determination, and the collision probability of the flying bird with the wind turbine blade is determined.

[0071] The present invention can effectively identify flying birds around wind turbines and analyze the probability of the flying birds' flight trajectories hitting the blades of wind turbines, which helps to reduce the load of the monitoring system and avoid disturbances to the monitoring system caused by other factors.

[0072] Example 3

[0073] Based on Example 2, the target tracking and comparison unit includes:

[0074] A pixel processing component is used to obtain the pixel value of each pixel point of a plurality of slice single frames of the initial video of a preset time length collected by the video acquisition unit, take the lower left corner of each slice single frame as the origin, store the pixel values ​​of all the pixel points on each slice single frame into a pixel array according to the coordinate values, and store the pixel arrays corresponding to the plurality of slice single frames into a linked list in chronological order;

[0075] The target tracking component is used to compare the pixel values ​​of the pixel points at the same coordinate position in the pixel arrays of adjacent slice single frames in sequence, store the coordinate values ​​of all the pixel points with pixel values ​​difference between each group of adjacent slice single frames in the position array, and store the position arrays of several groups of adjacent slice single frames in the position linked list in chronological order;

[0076] The target comparison component is used to extract the tracking target data features based on the target tracking component, and compare it with the pre-stored recorded bird posture feature data to determine whether the tracking target is a flying bird;

[0077] The trajectory extraction component is used to accumulate the X-axis coordinate value and the Y-axis coordinate value in each position array of the tracked target in the position linked list and divide them by the number of coordinate values ​​to obtain a number of target overall coordinate values, and store the several target overall coordinate values ​​in the overall position linked list in chronological order.

[0078] The beneficial effects of the above technical solution are as follows: after obtaining the pixel values ​​of all pixels of each sliced ​​single frame in the original video, taking the lower left corner of the sliced ​​single frame as the origin, assigning the coordinate value of the rectangular coordinate system to each pixel of the sliced ​​single frame, storing the coordinate values ​​and pixel values ​​of all pixels in the data group, realizing the digital recording of a single sliced ​​single frame, and then comparing the pixel values ​​of the pixels with the same coordinate values ​​in the tangent sliced ​​single frame. In this way, when there is a moving object in the sliced ​​single frame, only the pixel values ​​of the moving part will be different, so as to distinguish the pixel coordinate values ​​of the moving part, obtain the determination of the moving object stored in the sliced ​​single frame and the coordinate information of the moving object, on the basis of distinguishing the presence of the moving object in the sliced ​​single frame, extract the features of the coordinate information of the moving object and compare them with the flying bird posture feature data to determine whether the moving object is a flying bird, after the flying bird determination, the X-axis and Y-axis of the coordinates of the moving object determined to be a flying bird are accumulated and divided by the number of coordinate values ​​to obtain the coordinate value of the flying bird object, and the moving trajectory of the flying bird can be obtained by vectorizing a number of overall coordinate values;

[0079] The present invention can accurately determine the position of a moving object through the difference in pixel values ​​caused by the front and back position changes of the moving object, can effectively track the moving target, and can screen and determine the moving target, which can effectively avoid wasting resources on non-flying bird objects.

[0080] Example 4

[0081] Based on Example 1, the monitoring and early warning module includes:

[0082] A wind turbine impact detection and alarm unit is used to obtain the real-time impact situation of the wind turbine within a preset monitoring period based on the wind turbine impact matrix, and determine whether an early warning reminder of the impact is required based on the real-time impact situation;

[0083] The wind turbine blade damage detection and alarm unit is used to evaluate the impact damage of the wind turbine blades within a preset monitoring period, and determine whether to alarm based on the impact damage of the wind turbine blades.

[0084] Preferably, the wind turbine collision detection alarm unit comprises:

[0085] The vibration frequency collection module is used to collect the vibration frequency generated by the collision between birds and blades during the working state of the wind turbine in each preset monitoring period, and classify the frequency level according to the vibration frequency characteristics;

[0086] A vibration frequency monitoring matrix building module is used to build a vibration frequency monitoring matrix A of the entire blade of the wind turbine generator based on the vibration frequency generated when birds collide with the blades during the working state of the wind turbine generator in each preset monitoring period;

[0087] The impact trace collection module is used to collect the traces of birds hitting the blades during the working state of the wind turbine in each preset monitoring period, and classify the trace levels according to the characteristics of the traces;

[0088] The impact trace monitoring matrix building module is used to build a trace monitoring matrix B of the wind turbine blades based on the traces produced by the collision between birds and blades during the working state of the wind turbine in each monitoring cycle;

[0089] A wind turbine impact monitoring matrix construction module is used to construct a wind turbine blade impact monitoring matrix C based on a wind turbine blade vibration frequency monitoring matrix A and a wind turbine blade trace monitoring matrix B within each preset monitoring period;

[0090] The alarm evaluation module is used to evaluate whether an impact alarm is needed based on the evaluation index of the impact monitoring matrix C of the wind turbine blades in each preset monitoring cycle.

[0091] The beneficial effects of the above technical solution are as follows: in the impact detection alarm unit of the wind turbine generator, the vibration frequencies of various parts of the wind turbine generator blades are collected by the vibration frequency collection module, and the frequency levels are classified according to the vibration frequency characteristics, and the vibration frequency monitoring matrix A is constructed by the vibration frequency monitoring matrix construction module. At the same time, the impact trace collection module collects traces from various parts of the wind turbine generator blades, and the trace levels are classified according to the trace characteristics, and the impact trace monitoring matrix B is constructed by the impact trace monitoring matrix construction module. On the basis of the vibration frequency monitoring matrix A and the trace monitoring matrix B, an impact monitoring matrix C is integrated to form, and the impact of the wind turbine generator blades is monitored through the impact monitoring matrix C to determine whether the impact of the wind turbine generator blades needs to be alarmed;

[0092] Through comprehensive analysis of impact vibration frequency and impact marks, the collision of wind turbine blades can be accurately monitored and early warning of wind turbine blade collisions can be given in a timely manner, ensuring the response speed and accuracy of the wind turbine monitoring system and effectively extending the service life of wind turbine blades.

[0093] Example 5

[0094] On the basis of Example 4, based on the vibration frequency generated when birds collide with blades during the working state of the wind turbine in each preset monitoring period, a vibration frequency monitoring matrix A of the wind turbine blades is constructed:

[0095] Among them, A t is the vibration frequency monitoring matrix of the wind turbine blades in the tth preset monitoring period, a 11 is the vibration frequency of the wind turbine blade 11 area, a 12 is the vibration frequency of the wind turbine blade 12 area, a 1i is the vibration frequency of the wind turbine blade 1i area, a 21 is the vibration frequency of the wind turbine blade 21 area, a 22 is the vibration frequency of the wind turbine blade 22 area, a 2i is the vibration frequency of the wind turbine blade 2i area, a 31 is the vibration frequency of the wind turbine blade 31 area, a 32 is the vibration frequency of the wind turbine blade 32 area, a 3i is the vibration frequency of the wind turbine blade 3i area, a 41 is the vibration frequency of the wind turbine blade 41 region, a 42 is the vibration frequency of the wind turbine blade 42 region, a 4i is the vibration frequency of the wind turbine blade 4i area, a i1 is the vibration frequency of the wind turbine blade i1 region, ai2 is the vibration frequency of the i2 region of the wind turbine blade, a ii is the vibration frequency of the wind turbine blade region ii;

[0096] Based on the traces of birds colliding with blades during the working state of wind turbines, the impact trace monitoring matrix B of wind turbine blades is constructed:

[0097] Among them, B t is the collision trace monitoring matrix of the wind turbine blades in the tth preset monitoring period, b 11 is the impact mark characteristic of the wind turbine blade 11 area, b 12 is the impact mark characteristic of the wind turbine blade 12 area, b 1j b is the impact mark feature of the wind turbine blade 1j area, b 21 is the impact mark characteristic of the wind turbine blade 21 area, b 22 is the impact mark characteristic of the wind turbine blade 22 area, b 2j b is the impact mark feature of the wind turbine blade 2j area, b 31 is the impact mark characteristic of the wind turbine blade 31 area, b 32 is the impact mark characteristic of the wind turbine blade 32 area, b 3j b is the impact mark feature of the 3j area of ​​the wind turbine blade, 41 is the impact mark characteristic of the wind turbine blade 41 area, b 42 is the impact mark characteristic of the wind turbine blade 42 area, b 4j b is the impact mark feature of the 4j area of ​​the wind turbine blade, j1 is the impact mark feature of the wind turbine blade j1 area, b j2 is the impact mark feature of the wind turbine blade j2 area, b jj It is the impact mark characteristic of the jj area of ​​the wind turbine blade.

[0098] The beneficial effects of the above technical solution are: by constructing the vibration frequency monitoring matrix A and the collision mark monitoring matrix B to integrate the collision monitoring matrix C, the intensity of the collision is monitored, and the collision of the wind turbine blades is warned in time, avoiding the risk of not being able to detect the collision of the wind turbine blades in time.

[0099] Example 6

[0100] On the basis of Example 5, the impact monitoring matrix C of the wind turbine blades is constructed based on the vibration frequency monitoring matrix A of the wind turbine blades and the impact mark monitoring matrix B of the wind turbine blades:

[0101] Among them, C t is the impact monitoring matrix of the wind turbine blade in the tth preset monitoring period, λ is the matching weight of the vibration frequency of the wind turbine blade, and τ is the matching weight of the impact mark of the wind turbine blade;

[0102] When C t When the value of any element in the collision monitoring matrix exceeds the preset rated value, C t The subscript information and value of the element in the collision monitoring matrix are output and an early warning prompt is given.

[0103] The beneficial effects of the above technical solution are: constructing a collision monitoring matrix C based on the vibration frequency monitoring matrix A and the collision mark monitoring matrix B, intuitively understanding the impact situation and impact position of the wind turbine blades through the dynamic numerical changes of the vibration frequency and the collision mark, automatically determining the impact level and issuing an early warning, improving the management level of the wind turbine blades, and increasing the power generation efficiency of a single wind turbine blade.

[0104] Example 7

[0105] Based on Example 4, the wind turbine blade damage detection and alarm unit includes:

[0106] The blade damage capture module is used to compare the real-time contour data of the bird impact position of the wind turbine blade with the original contour data based on the bird impact position determined by the wind turbine impact detection alarm unit. If the real-time contour data and the original contour data do not match, it is preliminarily determined that the blade has been damaged by impact and the coordinate data of the mismatched position is captured;

[0107] The blade damage confirmation module is used to extract the pixels and features of the damaged parts of the blade based on the impact damage location coordinates determined by the blade damage capture module, and input them into the damage resolution model trained with various damage feature data to judge and screen the blade damage.

[0108] In this embodiment, the damage identification model is a model that is trained with big data by taking damage feature data of a number of wind turbine blades as input and taking comparison and determination of the feature damage data as output.

[0109] The beneficial effects of the above technical solution are: by real-time monitoring of the wind turbine blades, the blade impact marks are captured, and by comparing the impact mark with the original blade contour, it is determined whether the blade is damaged. After preliminarily determining that the blade is damaged, the pixels and features of the damaged area are extracted and input into the damage resolution model to identify and determine the type of damage. The present invention can monitor the damaged position of the wind turbine blades in real time, and based on the judgment of the damage, it can enable maintenance personnel to understand the situation in a timely manner and perform maintenance in a timely manner according to the situation.

[0110] Example 8

[0111] On the basis of Example 7, it also includes:

[0112] The blade damage assessment module is used to calculate the area of ​​collision damage and assess the hazard level of collision damage to the blade based on the type judgment and feature data of the collision damage confirmed by the blade damage confirmation module;

[0113] The damage expansion calculation module is used to evaluate the damage expansion coefficient of the current collision damage of the blade when the wind turbine continues to work based on the collision damage area and collision damage hazard level of the blade damage assessment module;

[0114] Preferably, the alarm level calculation unit is used to obtain the current blade damage alarm level based on the damage expansion coefficient evaluated by the blade performance damage calculation module.

[0115] Preferably, the damage expansion coefficient of the current wind turbine blade is:

[0116] Among them, E d is the damage expansion coefficient of the wind turbine blade at the current moment, ρ is the air density, A is the blade collision damage area, v is the blade rotation speed, K is the action coefficient, ε is the blade collision damage drag coefficient, μ is the mass conversion coefficient, and S is the blade collision damage circumference;

[0117] Current blade damage alarm level:

[0118] Among them, Β is the damage alarm level of the wind turbine blade at the current moment, o is the damage basic coefficient, and Δ is the collision damage hazard level.

[0119] The beneficial effects of the above technical solution are: by calculating the damage area of ​​the wind turbine blade, evaluating the hazard level of the collision damage to the blade, and calculating the expansion coefficient of the damaged part of the wind turbine blade when it continues to work based on the area and hazard level of the collision damage, judging the alarm level of the current blade collision damage based on the expansion coefficient, and then judging what kind of subsequent treatment the blade needs. The present invention can manage the blade damage in a precise manner, and can dynamically adjust the maintenance plan according to the blade damage, which can not only avoid further aggravation of the damage after the blade is damaged, but also reduce unnecessary shutdown of the wind turbine.

[0120] The present invention is subject to various modifications and variations without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A monitoring system for a wind turbine, characterized in that: include: A feature acquisition module, used to collect videos around wind turbines; The data storage and comparison module is used to store and record the bird posture data in advance, compare and match the feature acquisition module data through the trained big data comparison and matching model, extract the bird trajectory in the feature acquisition module data after identification and matching, and confirm the collision probability; The monitoring and early warning module is used to build a wind turbine impact damage monitoring matrix, collect impact data based on the bird impact probability of the data storage comparison module, and determine whether the wind turbine needs an early warning according to the impact monitoring matrix. At the same time, it marks the impact area, continuously monitors the impact area, and determines whether to report a damage warning based on the characteristics of the impact area; The collaborative recording module is used to adjust the operating state of the wind turbine according to the instructions transmitted by the monitoring and early warning module, and store the monitoring records and instruction records of the monitoring and early warning module.

2. A monitoring system for a wind turbine according to claim 1, characterized in that: The feature data storage comparison module includes: A video frame slicing unit, used for repeatedly slicing the initial video of a preset time length acquired by the video acquisition unit frame by frame; A target tracking and comparison unit is used to input a number of slice single frames based on a preset time length into a comparison algorithm model to identify a moving target; A screening storage unit is used to pre-store and record the bird posture data, and temporarily store the target slice single frame that exists in the target tracking unit to identify the target, and delete the remaining target slice single frame that does not exist in the tracking unit to identify the target; The trajectory analysis unit is used to determine the flight trajectory of the bird based on the comparison and recognition data of the target tracking and comparison unit, and analyze the probability of the bird's flight trajectory hitting the wind turbine blades.

3. A monitoring system for a wind turbine according to claim 2, characterized in that: The target tracking and comparison unit includes: A pixel processing component is used to obtain the pixel value of each pixel point of a plurality of slice single frames of the initial video of a preset time length collected by the video acquisition unit, take the lower left corner of each slice single frame as the origin, store the pixel values ​​of all the pixel points on each slice single frame into a pixel array according to the coordinate values, and store the pixel arrays corresponding to the plurality of slice single frames into a linked list in chronological order; The target tracking component is used to compare the pixel values ​​of the pixel points at the same coordinate position in the pixel arrays of adjacent slice single frames in sequence, store the coordinate values ​​of all the pixel points with pixel values ​​difference between each group of adjacent slice single frames in the position array, and store the position arrays of several groups of adjacent slice single frames in the position linked list in chronological order; The target comparison component is used to extract the tracking target data features based on the target tracking component, and compare it with the pre-stored recorded bird posture feature data to determine whether the tracking target is a flying bird; The trajectory extraction component is used to accumulate the X-axis coordinate value and the Y-axis coordinate value in each position array of the tracked target in the position linked list and divide them by the number of coordinate values ​​to obtain a number of target overall coordinate values, and store the several target overall coordinate values ​​in the overall position linked list in chronological order.

4. A monitoring system for a wind turbine according to claim 1, characterized in that: The monitoring and early warning modules include: A wind turbine impact detection and alarm unit is used to obtain the real-time impact situation of the wind turbine within a preset monitoring period based on the wind turbine impact matrix, and determine whether an early warning reminder of the impact is required based on the real-time impact situation; The wind turbine blade damage detection and alarm unit is used to evaluate the impact damage of the wind turbine blades within a preset monitoring period, and determine whether to alarm based on the impact damage of the wind turbine blades.

5. A monitoring system for a wind turbine according to claim 4, characterized in that: The wind turbine impact detection alarm unit includes: The vibration frequency collection module is used to collect the vibration frequency generated by the collision between birds and blades during the working state of the wind turbine in each preset monitoring period, and classify the frequency level according to the vibration frequency characteristics; A vibration frequency monitoring matrix building module is used to build a vibration frequency monitoring matrix A of the entire blade of the wind turbine generator based on the vibration frequency generated when birds collide with the blades during the working state of the wind turbine generator in each preset monitoring period; The impact trace collection module is used to collect the traces of birds hitting the blades during the working state of the wind turbine in each preset monitoring period, and classify the trace levels according to the characteristics of the traces; The impact trace monitoring matrix building module is used to build a trace monitoring matrix B of the wind turbine blades based on the traces produced by the collision between birds and blades during the working state of the wind turbine in each monitoring cycle; A wind turbine impact monitoring matrix construction module is used to construct a wind turbine blade impact monitoring matrix C based on a wind turbine blade vibration frequency monitoring matrix A and a wind turbine blade trace monitoring matrix B within each preset monitoring period; The alarm evaluation module is used to evaluate whether an impact alarm is needed based on the evaluation index of the impact monitoring matrix C of the wind turbine blades in each preset monitoring cycle.

6. A monitoring system for a wind turbine according to claim 5, characterized in that: Based on the vibration frequency generated when birds collide with blades during the working state of the wind turbine in each preset monitoring period, the vibration frequency monitoring matrix A of the wind turbine blades is constructed: Among them, A t is the vibration frequency monitoring matrix of the wind turbine blades in the tth preset monitoring period, a 11 is the vibration frequency of the wind turbine blade 11 area, a 12 is the vibration frequency of the wind turbine blade 12 area, a 1i is the vibration frequency of the wind turbine blade 1i area, a 21 is the vibration frequency of the wind turbine blade 21 area, a 22 is the vibration frequency of the wind turbine blade 22 area, a 2i is the vibration frequency of the wind turbine blade 2i area, a 31 is the vibration frequency of the wind turbine blade 31 area, a 32 is the vibration frequency of the wind turbine blade 32 area, a 3i is the vibration frequency of the wind turbine blade 3i area, a 41 is the vibration frequency of the wind turbine blade 41 region, a 42 is the vibration frequency of the wind turbine blade 42 region, a 4i is the vibration frequency of the wind turbine blade 4i area, a i1 is the vibration frequency of the wind turbine blade i1 region, a i2 is the vibration frequency of the i2 region of the wind turbine blade, a ii is the vibration frequency of the wind turbine blade region ii; Based on the traces of birds colliding with blades during the working state of wind turbines, the impact trace monitoring matrix B of wind turbine blades is constructed: Among them, B t is the vibration frequency monitoring matrix of the wind turbine blades in the tth preset monitoring period, b 11 is the impact mark characteristic of the wind turbine blade 11 area, b 12 is the impact mark characteristic of the wind turbine blade 12 area, b 1j b is the impact mark feature of the wind turbine blade 1j area, b 21 is the impact mark characteristic of the wind turbine blade 21 area, b 22 is the impact mark characteristic of the wind turbine blade 22 area, b 2j b is the impact mark feature of the wind turbine blade 2j area, b 31 is the impact mark characteristic of the wind turbine blade 31 area, b 32 is the impact mark characteristic of the wind turbine blade 32 area, b 3j b is the impact mark feature of the 3j area of ​​the wind turbine blade, 41 is the impact mark characteristic of the wind turbine blade 41 area, b 42 is the impact mark characteristic of the wind turbine blade 42 area, b 4j b is the impact mark feature of the 4j area of ​​the wind turbine blade, j1 is the impact mark feature of the wind turbine blade j1 area, b j2 is the impact mark feature of the wind turbine blade j2 area, b jj It is the impact mark characteristic of the jj area of ​​the wind turbine blade.

7. A monitoring system for a wind turbine according to claim 6, characterized in that: Based on the vibration frequency monitoring matrix A of the wind turbine blades and the impact trace monitoring matrix B of the wind turbine blades, the impact monitoring matrix C of the wind turbine blades is constructed: Among them, c t is the impact monitoring matrix of the wind turbine blade in the tth preset monitoring period, λ is the matching weight of the vibration frequency of the wind turbine blade, and τ is the matching weight of the impact mark of the wind turbine blade; When C t When the value of any element in the collision monitoring matrix exceeds the preset rated value, C t The subscript information and value of the element in the collision monitoring matrix are output and an early warning prompt is given.

8. A monitoring system for a wind turbine according to claim 4, characterized in that: The wind turbine blade damage detection and alarm unit includes: The blade damage capture module is used to compare the real-time contour data of the bird impact position of the wind turbine blade with the original contour data based on the bird impact position determined by the wind turbine impact detection alarm unit. If the real-time contour data and the original contour data do not match, it is preliminarily determined that the blade has been damaged by impact and the coordinate data of the mismatched position is captured; The blade damage confirmation module is used to extract the pixels and features of the damaged parts of the blade based on the impact damage location coordinates determined by the blade damage capture module, and input them into the damage resolution model trained with various damage feature data to judge and screen the blade damage.

9. A monitoring system for a wind turbine according to claim 8, characterized in that: Also includes: The blade damage assessment module is used to calculate the area of ​​collision damage and assess the hazard level of collision damage to the blade based on the type judgment and feature data of the collision damage confirmed by the blade damage confirmation module; The damage expansion calculation module is used to evaluate the damage expansion coefficient of the current collision damage of the blade when the wind turbine continues to work based on the collision damage area and collision damage hazard level of the blade damage assessment module; The alarm level calculation unit is used to obtain the current blade damage alarm level based on the damage expansion coefficient evaluated by the blade performance damage calculation module.

10. A monitoring system for a wind turbine according to claim 9, characterized in that: Current damage expansion factor of wind turbine blades: Among them, E d is the damage expansion coefficient of the wind turbine blade at the current moment, ρ is the air density, A is the blade collision damage area, v is the blade rotation speed, K is the action coefficient, ε is the blade collision damage drag coefficient, μ is the mass conversion coefficient, and S is the blade collision damage circumference; Current blade damage alarm level: Among them, Β is the damage alarm level of the wind turbine blade at the current moment, o is the damage basic coefficient, and Δ is the collision damage hazard level.

Images