Method and system for optimizing clearance value based on three-beam spatial constraint

Abstract

The application provides a three-beam space constraint-based clearance value optimization method and system, and relates to the technical field of laser radar. The three-beam space constraint-based clearance value optimization method comprises the following steps: a clearance radar collects monitoring data of three groups of beams, and the collected monitoring data comprises a time sequence stamp, a light return intensity and a ranging value; the system pre-processes the collected data, calculates a clearance value after eliminating abnormal data of the light return intensity and the ranging value; a common reasonableness constraint of a blade trigger is set; a trigger scene is divided by determining a trigger state of the three groups of beams, and a trigger scene corresponding to a moment is identified; for different trigger scenes, a reasonableness interval of the clearance value of the three groups of beams is set; all data at the moment meet the constraint and an experience rule, and an effective clearance value is output.

Description

Technical Field

[0001] This application belongs to the field of lidar technology, and in particular relates to a method and system for optimizing the clearance value based on three-beam spatial constraints. Background Technology

[0002] Current wind turbine blade clearance monitoring largely relies on simple multi-beam ranging with laser clearance radar, outputting clearance values ​​through basic geometric calculations. Under complex operating conditions, interference from rain, snow, fog, blade deformation, and signal noise can easily lead to problems such as misjudgment of clearance values, poor data reliability, and insufficient algorithm robustness, making accurate and reliable clearance monitoring difficult. Traditional clearance optimization algorithms lack systematic constraint mechanisms, fail to identify different scenarios, and do not establish clear reasonable ranges. Clearance judgments often rely on empirical values, simply setting a single upper or lower limit, or relying solely on simple relationships between three sets of beams for verification. This makes it difficult to adapt to the dynamic operating requirements of wind turbines and cannot effectively avoid the risk of blade-tower collisions.

[0003] To address the aforementioned issues, there is an urgent need for a method to optimize the net clearance value based on beam space constraints, incorporating explicit empirical rules, adaptively compensating for blade deformation, and possessing precise geometric modeling capabilities. This method would enable accurate verification and reliable output of the wind turbine blade net clearance value, ensuring the long-term safe and stable operation of the wind turbine.

[0004] Existing laser-based air clearance radar monitoring systems primarily acquire ranging data by emitting multiple probe laser beams and receiving the scattered echoes from the blades. This data is then combined with basic geometric formulas to calculate the blade clearance value, providing data support for wind turbine safety control. The basic principle is to calculate the horizontal distance from the blade tip to the outer edge of the tower based on radar ranging data and tower foundation parameters, thus characterizing the blade clearance status. However, traditional algorithms lack a robust constraint mechanism, fail to construct a scenario-specific recognition system, and do not define a clear reasonable range for clearance values. Clearance verification often relies on manual experience, simply setting a single upper or lower limit, or judging based solely on simple numerical relationships between three sets of beams, without conducting detailed geometric derivations or considering the impact of blade deformation. This makes it difficult to meet the clearance monitoring needs under complex operating conditions. The existing technology has the following problems: 1. Lack of constraint mechanism: Traditional algorithms lack a systematic constraint system and have not built a two-layer reasonable constraint model. The clearance judgment relies on human experience and lacks standardized and quantitative constraint standards, resulting in low verification accuracy; 2. Gap in scene recognition: No trigger scene recognition system has been established, and multiple trigger combinations are not covered. It is impossible to distinguish between normal and abnormal trigger states, resulting in poor adaptability; 3. Ambiguous reasonable range: The reasonable range of clearance value is not clearly defined. Only a single upper or lower limit is simply set, or a simple relationship between three sets of beams is relied upon. The range definition is inaccurate and incomplete. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art. The present invention proposes a method and system for optimizing the clearance value of lidar based on three sets of beam spatial constraints. By constructing a two-layer rationality constraint model and integrating an empirical rule verification mechanism, the accurate optimization and reliable output of the clearance value of wind turbine blades can be achieved.

[0006] The technical solution of the present invention is as follows: On the one hand, the present invention discloses a method for optimizing the clearance value based on three-beam spatial constraints. The method includes the following steps: S1, the clearance radar collects monitoring data of three sets of echo beams. The collected monitoring data includes time stamp, echo intensity and ranging value. S2. The system preprocesses the collected data, removes abnormal data in backlight intensity and distance measurement, and then calculates the clearance value. S3. Set common rationality constraints for blade triggering: Based on the effective blade length, define an initial rationality interval based on the distance threshold from the target point to the blade root, and collect the wind speed at the current moment in real time. Fan speed Pitch angle Through preset mapping relationships The nonlinear deformation of the blade was calculated. ,based on The initial reasonableness range is adaptively adjusted, and Incorporate constraint models; S4. The triggering scene is divided by determining the triggering state of the three sets of beams, and the triggering scene at the corresponding time is identified. S5. For different triggering scenarios, based on the spatial geometric relationship between the beam and the blade, construct the blade straight line equation and the beam straight line equation, solve the intersection point simultaneously to obtain the upper and lower limits of the beam clearance value, and then combine the values ​​calculated in real time in step S3. Substitute the reasonable range of the clearance value, compensate for the lower limit of the range, and set reasonable ranges for the clearance values ​​of the three beams respectively; S6. At this moment, all data satisfy the constraints and empirical rules, and the effective net clearance value is output.

[0007] S2 includes the following steps: S2.1, Preset threshold I based on backlight intensity th Remove low signal-to-noise ratio data, remove invalid ranging values ​​affected by weather, and mark the validity status of the three sets of beam data; S2.2. Based on radar geometry, obtain the real-time clearance value of beam n (n=1,2,3). The formula is expressed as: ; in, The effective ranging value of beam n Let n be the angle between the beam and the vertical direction. The horizontal distance between the radar and the center of the tower. The equivalent radius of the blade tip sweeping across the plane of the tower. For beam The horizontal component.

[0008] In S3, the common rationality constraints for blade triggering include blade ranging range constraints and deformation adaptive correction. Among the blade ranging range constraints, when the beam triggers the blade, the distance L from the target point to the blade root must satisfy the following formula: ; in, The effective length of the blade. , The coordinates of the target point, The coordinates of the leaf root are... The nonlinear deformation of the blades is determined by wind speed. Rotation speed Pitch angle The deformation of the blade tip towards the tower is calculated in real time and used to correct the effective detection range of the blade tip and compensate for the positional offset caused by the bending of the blade. The deformation adaptive correction is used to adaptively adjust the reasonable range of the net clearance value based on the real-time calculated ΔL.

[0009] In S4, with beam 3 as the priority core, the triggering scenarios are divided into the first scenario, the second scenario, the third scenario, and the fourth scenario. In the first scenario, beam 3 does not trigger the blades; in the second scenario, only beam 3 triggers the blades; in the third scenario, beams 2 and 3 trigger the blades, while beam 1 does not trigger; in the fourth scenario, beams 1, 2, and 3 all trigger the blades.

[0010] S5 includes the following steps: S5.1 Linear Constraint Model Construction: Based on the included angle of the three beams and the blade's geometric confidentiality parameters, the blade's linear equation and the beam's linear equation are constructed. Solving these equations simultaneously yields the upper and lower limits of the three beams' clearance values. The upper limit constraint formula for the clearance value, where the target beam m hits the blade tip, is as follows: ; in, The vertical height from the blade tip to the radar mounting surface; Let m be the angle between the beam and the vertical direction; When another beam n hits the blade tip, the equation of the blade's straight line is: ; in, Leaf slope , The x-coordinate is the root of the leaf. The vertical coordinate is the base of the leaf. Let x be the x-coordinate of the blade tip point triggered by beam n. Let be the ordinate of the blade tip point triggered by beam n; The equation of the straight line for beam m is: ; By combining the linear equations of the blade and the beam m, the x-coordinate of the intersection point can be obtained. We can obtain: ; Substituting the x-coordinate of the intersection point into the clearance value formula, we can obtain the lower clearance value of beam m as follows: ; in, The x-coordinate of the intersection point of the blade line and the beam line m; When beams m and n fail to hit the blade tip, it is necessary to solve the equations for the blade and the beam simultaneously to calculate the upper limit of the clearance value corresponding to the beam. Lower limit of net air value ; S5.2 Blade Deformation Constraint Superposition: The real-time calculated nonlinear blade deformation variable ΔL is substituted into the rationality interval, and adaptive compensation is performed on the interval. The rationality interval after compensation is: ; S5.3 Verification: At this moment, all three beams fall within their respective reasonable ranges, satisfying the single-beam spatial constraint requirements.

[0011] In S6, the constraints and empirical rules include spatial association rules, temporal gradient rules, data validity rules, and numerical lower limit rules. In the spatial association rules, if beam 1 is triggered, beams 2 and 3 must also be triggered, and if beam 2 is triggered, beam 3 must also be triggered. In the temporal gradient rules, the scene at the current moment and the scenes at the previous and next moments are adjacent level jumps, and the clearance value is only allowed to jump to the adjacent scene, and cross-scene abrupt changes are prohibited. In the data validity rules, all triggered beam data are valid data, and invalid values ​​are prohibited. In the numerical lower limit rules, the clearance value of a single beam must not be lower than the minimum clearance value when all three sets of beams are triggered.

[0012] In S6, the effective clearance value follows the minimum function principle, selecting the minimum clearance value of the effective trigger beam as the final reliable clearance value. The formula is: ; in, The clearance value for beam 2. This is the clearance value for beam 3.

[0013] In S1, the laser air clearance radar emits a detection laser beam toward the wind turbine blades to actively detect the air clearance status of the blades. During the signal propagation and scattering process, the air clearance radar collects data from the received echoes.

[0014] On the other hand, this invention discloses a clearance value optimization system, which includes a data acquisition module, a data preprocessing module, a constraint determination module, a scene recognition module, a constraint configuration module, and a clearance value calculation module. The data acquisition module is used to monitor and acquire data from three sets of beams. The data preprocessing module is used to preprocess the acquired data and perform preliminary calculations on the clearance value. The constraint determination module is used to verify the clearance value according to preset constraint conditions. The scene recognition module is used to identify the triggering scene after determining the triggering state of the beam. The constraint configuration module is used to configure the legal and reasonable value ranges corresponding to the real-time clearance values ​​of the three sets of beams respectively. The clearance value calculation module is used to output the final clearance value after integrating empirical rules. Attached Figure Description

[0015] Figure 1 This is a flowchart of the method of the present invention; Figure 2 This is a block diagram of the airspace radar safety monitoring interaction of the present invention; Figure 3 This is a simplified flowchart of the rationality optimization algorithm of the present invention; Figure 4 It is a scatter plot of the clearance values ​​of the three beams of the clearance radar and a trigger scenario diagram. Detailed Implementation

[0016] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0017] like Figures 1 to 4 As shown, on one hand, the present invention is a clearance value optimization method based on three-beam spatial constraints, which includes the following steps: S1. Acquisition of raw data from the airspace radar: The airspace radar acquires monitoring data of the three beams of the echo. The acquired monitoring data includes the time stamp, echo intensity, and distance measurement value. S2. Data preprocessing and initial calculation of clearance value: The system preprocesses the collected data, removes abnormal data of backlight intensity and distance measurement value, and then calculates the clearance value. S3. Global Common Constraint Configuration: Set common rationality constraints triggered by the blades. Based on the effective blade length, define the initial rationality interval based on the distance threshold from the target point to the blade root, and collect the wind speed at the current moment in real time. Fan speed Pitch angle Through preset mapping relationships The nonlinear deformation of the blade was calculated. ,based on The initial reasonableness range is adaptively adjusted, and Incorporate constraint models; S4. Three-beam trigger scene recognition: The trigger scene is divided by judging the trigger state of the three sets of beams, and the trigger scene at the corresponding time is identified. S5. Single-beam spatial rationality constraint verification: For different triggering scenarios, based on the spatial geometric relationship between the beam and the blade, the blade straight line equation and the beam straight line equation are constructed, and the intersection point is solved simultaneously to obtain the upper and lower limits of the beam clearance value. Then, the values ​​calculated in real time in step S3 are used to verify the beam clearance value. Substitute the reasonable range of the clearance value, compensate for the lower limit of the range, and set reasonable ranges for the clearance values ​​of the three beams respectively; S6. Empirical rule fusion and net clearance value output: All data at this moment satisfy the constraints and empirical rules, and output the effective net clearance value.

[0018] S2 includes the following steps: S2.1, Preset threshold I based on backlight intensity th Remove low signal-to-noise ratio data, remove invalid ranging values ​​affected by weather, and mark the validity status of the three sets of beam data; S2.2. Based on radar geometry, obtain the real-time clearance value of beam n (n=1,2,3). The formula is expressed as: ; in, The effective ranging value of beam n Let n be the angle between the beam and the vertical direction. The horizontal distance between the radar and the center of the tower. The equivalent radius of the blade tip sweeping across the plane of the tower. For beam The horizontal component.

[0019] In the above formula, the following condition is satisfied: , ( (The angle between adjacent beams) For beam The horizontal component of the target point is used to calculate the horizontal distance from the target point to the outer edge of the tower, which is the clearance value.

[0020] In S3, the common rationality constraints for blade triggering include blade ranging range constraints and deformation adaptive correction. When the beam triggers the blade, the target point corresponding to the ranging value must be located within the effective range of 1 / 3 of the blade length to the blade tip after considering deformation correction. Therefore, in the blade ranging range constraints, when the beam triggers the blade, the distance L from the target point to the blade root must satisfy the following formula: ; in, The effective length of the blade. , The coordinates of the target point, The coordinates of the leaf root are... The nonlinear deformation of the blades is determined by wind speed. Rotation speed Pitch angle The deformation of the blade tip towards the tower is calculated in real time and used to correct the effective detection range of the blade tip and compensate for the positional offset caused by the bending of the blade. The deformation adaptive correction is used to adaptively adjust the reasonable range of the net clearance value based on the real-time calculated ΔL.

[0021] In S4, with beam 3 as the priority core, the triggering scenarios are divided into the first scenario, the second scenario, the third scenario, and the fourth scenario. In the first scenario, beam 3 does not trigger the blades; in the second scenario, only beam 3 triggers the blades; in the third scenario, beams 2 and 3 trigger the blades, while beam 1 does not trigger; in the fourth scenario, beams 1, 2, and 3 all trigger the blades.

[0022] S5 includes the following steps: S5.1 Linear Constraint Model Construction: Based on the included angle of the three beams and the blade's geometric confidentiality parameters, the blade's linear equation and the beam's linear equation are constructed. Solving these equations simultaneously yields the upper and lower limits of the three beams' clearance values. The upper limit constraint formula for the clearance value, where the target beam m hits the blade tip, is as follows: ; in, The vertical height from the blade tip to the radar mounting surface; Let m be the angle between the beam and the vertical direction; When another beam n hits the blade tip, the equation of the blade's straight line is: ; in, Leaf slope , The x-coordinate is the root of the leaf. The vertical coordinate is the base of the leaf. Let x be the x-coordinate of the blade tip point triggered by beam n. Let be the ordinate of the blade tip point triggered by beam n; The equation of the straight line for beam m is: ; By combining the linear equations of the blade and the beam m, the x-coordinate of the intersection point can be obtained. We can obtain: ; Substituting the x-coordinate of the intersection point into the clearance value formula, we can obtain the lower clearance value of beam m as follows: ; in, The x-coordinate of the intersection point of the blade line and the beam line m; When beams m and n fail to hit the blade tip, it is necessary to solve the equations for the blade and the beam simultaneously to calculate the upper limit of the clearance value corresponding to the beam. Lower limit of net air value ; S5.2 Blade Deformation Constraint Superposition: The real-time calculated nonlinear blade deformation variable ΔL is substituted into the rationality interval, and adaptive compensation is performed on the interval. The rationality interval after compensation is: ; S5.3 Verification: At this moment, all three beams fall within their respective reasonable ranges, satisfying the single-beam spatial constraint requirements.

[0023] In S6, the constraints and empirical rules include spatial association rules, temporal gradient rules, data validity rules, and numerical lower limit rules. In the spatial association rules, if beam 1 is triggered, beams 2 and 3 must also be triggered; if beam 2 is triggered, beam 3 must also be triggered. In the temporal gradient rules, the scene at the current moment and the scenes at the preceding and following moments must be adjacent-level transitions; the clearance value is only allowed to transition between adjacent scenes, and cross-scene abrupt changes are prohibited. In the data validity rules, all triggered beam data must be valid, and invalid values ​​are prohibited. In the numerical lower limit rules, the clearance value of a single beam must not be lower than the minimum clearance value when all three beams are triggered. In S6, the valid clearance value follows the minimum function principle, selecting the minimum clearance value of the valid triggered beams as the final reliable clearance value. The formula is: ; in, The clearance value for beam 2. This is the clearance value for beam 3.

[0024] In S1, the laser air clearance radar emits a detection laser beam toward the wind turbine blades to actively detect the air clearance status of the blades. During the signal propagation and scattering process, the air clearance radar collects data from the received echoes.

[0025] On the other hand, this invention discloses a clearance value optimization system applied to a clearance value optimization method based on three-beam spatial constraints. The system includes a data acquisition module, a data preprocessing module, a constraint determination module, a scene recognition module, a constraint configuration module, and a clearance value calculation module. The data acquisition module is used to monitor and acquire data from three sets of beams. The data preprocessing module is used to preprocess the acquired data and perform preliminary calculations on the clearance value. The constraint determination module is used to verify the clearance value according to preset constraint conditions. The scene recognition module is used to identify the triggering scene after determining the triggering state of the beam. The constraint configuration module is used to configure the legal and reasonable value ranges corresponding to the real-time clearance values ​​of the three sets of beams respectively. The clearance value calculation module is used to output the final clearance value after integrating empirical rules.

[0026] like Figure 2 As shown, multiple detection laser beams (beam 1, beam 2, and beam 3) are emitted towards the wind turbine blades via a laser-guided air clearance radar to actively detect the blade clearance status. During signal propagation and scattering, the echoes are received and processed to acquire the time stamp, echo intensity, and original ranging data corresponding to each beam. Based on the acquired raw data, the system sequentially performs preprocessing, dual-layer rationality constraint verification, and empirical rule fusion optimization, outputting a reliable clearance value. The optimized clearance value is input to the wind turbine control unit, which then generates a safety control strategy and applies it to the wind turbine.

[0027] Within the aforementioned interactive framework for airspace radar safety monitoring, to standardize the geometric relationships and calculation logic for subsequent airspace value calculation and constraint modeling, this invention first establishes a unified coordinate system and defines key parameters uniformly. The coordinate system takes the radar installation position as its origin O(0,0), and the horizontal axis is away from the tower center. The positive direction of the axis, perpendicular to the axis downwards is Positive direction of the axis.

[0028] During system operation, the laser air clearance radar collects raw data from the three-beam monitoring in real time at a high-frequency sampling rate, providing input support for subsequent processing. The collected raw data includes time stamps, three-beam ranging values, and three-beam backlight intensities. The definitions of each data point are as follows: Time Stamp Record the absolute timestamp of the data acquisition for subsequent time-series variation verification; beam ranging value. Real-time spatial distance measurements (unit: cm) for beams 1, 2, and 3, respectively; backlight intensity. The echo signal intensities (unit: dimensionless, range 0~100) of beams 1, 2, and 3 are used to determine signal validity and the degree of weather interference. The collected data is output to the data processing unit in real time, providing raw input for subsequent preprocessing, constraint verification, and clearance value optimization.

[0029] Figure 4 This is a scatter plot of the clearance values ​​of the three beams of the clearance radar, where blue dots represent the clearance value of beam 1 (S1), orange dots represent the clearance value of beam 2 (S2), and green dots represent the clearance value of beam 3 (S3). The horizontal axis represents the time index, and the vertical axis represents the clearance value. The system determines the triggering status (trigger blade / ground value) of each beam in real time. With beam 3 (S3) as the priority core, the triggering status of the three beams is divided into four standard scenarios. This embodiment focuses on demonstrating three typical scenarios of blade triggering: Second scenario (only S3 triggers blades): corresponding to the red box area at the top of the figure, at this time only the green S3 clearance value is valid, and there is no blade trigger data for blue S1 and orange S2. Scenario 3 (S2 and S3 trigger blades, S1 does not trigger): This corresponds to the red box area in the middle of the diagram. At this time, the orange S2 and green S3 clearance values ​​are valid and meet the following conditions: Spatial gradient relationship, blue S1 no-blade trigger data; Fourth scenario (S1, S2, and S3 all trigger the blade): corresponds to the red box area at the bottom of the diagram. At this time, the blue S1, orange S2, and green S3 clearance values ​​are all valid and display " The continuous gradient distribution characteristics of "". The first scenario (S3 ranging value is the ground reflection value, no blades are triggered) has no blade trigger clearance data and is not shown in this figure. The above scenario division is based on beam spatial priority logic, which highly corresponds to the clearance value distribution characteristics in the figure.

[0030] For different triggering scenarios, the system establishes a single-beam spatial rationality constraint model based on the spatial geometric relationship between the beam and the blade, and superimposes blade deformation redundancy to set the rationality range of the clearance value of each beam.

[0031] Taking the second scenario (only S3 triggers the blade) as an example, when beam 3 hits the blade tip, its upper limit of clearance can be expressed as: ; in, The vertical height from the blade tip to the radar mounting surface; The angle between beam 3 and the vertical direction.

[0032] When beam 2 hits the blade tip, first construct the straight-line equation between the blade and beam 3: The linear equation of the leaf is derived from the leaf root. The blade tip triggered by beam 2 ; Determined as: ; Among them, the slope of the blade ; The equation of the straight line for beam 3 is: ; Solving the above two equations simultaneously, we can obtain the x-coordinate of the intersection point. , can be obtained ; Substituting the x-coordinate of the intersection point into the clearance value formula, we can obtain the lower clearance value of beam 3 as follows: ; in, The x-coordinate is the intersection point of the blade's straight line and the beam's 3rd line. The nonlinear deformation of the blade is superimposed. Then, the reasonable range for S3 in the second scenario is: ; The same principle applies to single-beam constraints in other scenarios. A linear constraint model is constructed based on the triggering state and spatial geometric relationship of the corresponding beam, and deformation redundancy is superimposed.

[0033] A laser-guided radar is used to collect raw data from three-beam monitoring in real time. The collected data includes timestamps, three-beam ranging values, and three-beam echo intensity, providing raw input for subsequent processing. In one embodiment, taking the measured data at 16:35:28 on April 16, 2026 as an example, the collected raw data is as follows: Time stamp: 2026-04-16, 16:35:28; Beam 1: Distance measured 15958cm, backlight intensity 74; Beam 2: Distance measured 8471cm, backlight intensity 65; Beam 3: Distance measurement 7333cm, backlight intensity 85.

[0034] Preset threshold based on backlight intensity Low signal-to-noise ratio data are removed. At this moment, the return light intensity of all three beams is higher than the preset threshold, and there is no invalid data. At the same time, invalid ranging values ​​caused by exceeding the range or weather interference are removed. There is no such abnormal data at this moment, and the three beam data is marked as valid.

[0035] Based on the system's built-in security tower geometry parameters, the initial calculation formula for the clearance value is adopted: ; in, For beam The effective distance measurement value, For beam The angle with the vertical direction, The horizontal distance between the radar and the center of the tower. The equivalent radius of the blade tip sweeping plane tower (all the above parameters are built-in confidential values ​​of the system).

[0036] The net clearance value of the three beams at this moment is calculated as follows: Beam 1 clearance value cm; Beam 2 clearance value cm; Beam 3 clearance value cm.

[0037] When the beam triggers the blade, the distance L from the target point to the blade root must satisfy: ; in, This is the effective length of the blade (a built-in confidential value in the system). The nonlinear deformation of the blades is determined by wind speed. Rotation speed Pitch angle Real-time calculations are performed to compensate for the bending displacement of the blade tip towards the tower. At this moment, the target point distances corresponding to the ranging values ​​of beams 2 and 3 both fall within the effective range, satisfying the constraint. Beam 1 does not satisfy the constraint, and therefore represents the ground value.

[0038] The deformation adaptive correction adopts blade deformation redundancy constraints. Based on the real-time calculated ΔL, the reasonable range of the clearance value is adaptively adjusted. At this moment, deformation compensation has been incorporated into the subsequent verification logic.

[0039] The triggering status (triggered blade / untriggered blade) of each beam is determined in real time, and the triggering scenario at this moment is identified with beam 3 (S3) as the priority core: The target point corresponding to the ranging value of beam 1 does not fall within the effective triggering range of the blade, and is therefore determined to be a blade that has not been triggered; The target points corresponding to the ranging values ​​of beams 2 and 3 both fall within the effective triggering range of the blade, and are therefore determined to be triggering blades.

[0040] Therefore, at this moment, S2 and S3 trigger the blades, while S1 does not trigger, which perfectly matches the definition of the third scenario. This moment is the third scenario.

[0041] For the third scenario, based on the spatial geometric relationship between the beam and the blade, a linear constraint model is constructed and blade deformation redundancy is superimposed to set a reasonable range for the clearance values ​​of beam 2 and beam 3: Based on the system's built-in three-beam angle and blade geometric confidentiality parameters, the blade linear equation and the beam linear equation are constructed and solved simultaneously to obtain the upper and lower limit ranges of the clearance values ​​of beam 2 and beam 3; the real-time calculated blade nonlinear deformation ΔL is substituted into the reasonable range and adaptive compensation is performed on the range; at this moment, the clearance value of beam 2 (822cm) and the clearance value of beam 3 (948cm) both fall within their respective reasonable ranges, satisfying the single-beam spatial constraint requirements.

[0042] To enhance the reliability and interpretability of the algorithm, the system integrates four core empirical rules and establishes standardized clearance value output, abnormal value compensation, and judgment logic. These include: a spatial association rule (triggering beam 1 guarantees the triggering of beams 2 and 3, and vice versa); a temporal gradient rule (clearance values ​​are only allowed to change within adjacent scenes, prohibiting cross-scene changes); a data validity rule (invalid values ​​are prohibited in a set of clearance values); and a numerical lower limit rule (the clearance value of a single beam must not be lower than the minimum clearance value when all three beams are triggered).

[0043] In this embodiment, beam 3 is triggered synchronously when beam 2 is triggered, which conforms to the spatial association rule that "if beam 2 is triggered, beam 3 must be triggered"; the scene at this moment and the scenes at the previous and next moments are adjacent level jumps, without cross-scene abrupt changes, which conforms to the spatiotemporal gradation rule; the triggered beam data are all valid data with no invalid values, which conforms to the data validity rule; the clearance values ​​of the two triggered beams are not lower than the minimum clearance value when all three beams are triggered, and all verifications pass, which conforms to the numerical lower limit rule.

[0044] This embodiment follows the principle of minimum function, selecting the minimum clearance value of the effective trigger beam as the final reliable clearance value, as shown in the formula: Substitute the data, .

[0045] At this moment, all data meet the constraints and empirical rules, and the effective clearance value is output normally. Therefore, the effective clearance value in this embodiment is 822cm.

[0046] This invention is applicable to wind turbine safety monitoring in wind farms, laser airspace radar data processing, optimization of wind turbine blade clearance values, intelligent control of wind power equipment, and other laser radar systems requiring high-precision blade clearance monitoring. Based on the collected raw data, this invention performs preprocessing, dual-layer rationality constraint verification, and empirical rule fusion optimization to output a reliable clearance value. The optimized clearance value is input into the wind turbine control unit, which then generates a safety control strategy and applies it to the wind turbine, forming a closed-loop process of "detection-optimization-control".

[0047] This invention constructs a three-beam, two-layer rationality constraint model and a full-coverage trigger scene recognition system to clarify the rationality range of air clearance values. It integrates four core empirical rules and introduces blade nonlinear deformation compensation, conducting detailed geometric relationship modeling to achieve accurate verification and dynamic optimization of wind turbine blade air clearance values. This method possesses advantages such as a comprehensive constraint system, strong scene adaptability, accurate rationality range, sufficient deformation compensation, rigorous geometric modeling, and high algorithm robustness. It effectively avoids misjudgment of air clearance values ​​caused by weather interference, blade deformation, and signal noise, significantly improving the accuracy, stability, and reliability of laser air clearance radar monitoring data. This provides solid data support for wind turbine safety control and is applicable to wind farm turbine safety monitoring scenarios under various complex operating conditions.

[0048] Finally, it should be emphasized that the above description is not intended to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A clearance value optimization method based on three-beam spatial constraints, comprising the following steps: S1. The airspace radar collects monitoring data from the three sets of echo beams. The collected monitoring data includes the time stamp, echo intensity, and distance measurement value. S2. The system preprocesses the collected data, removes abnormal data in backlight intensity and distance measurement, and then calculates the clearance value. S3. Set common rationality constraints for blade triggering: Based on the effective blade length, define an initial rationality interval based on the distance threshold from the target point to the blade root, and collect the wind speed at the current moment in real time. Fan speed Pitch angle Through preset mapping relationships The nonlinear deformation of the blade was calculated. ,based on The initial reasonableness range is adaptively adjusted, and Incorporate constraint models; S4. The triggering scene is divided by determining the triggering state of the three sets of beams, and the triggering scene at the corresponding time is identified. S5. For different triggering scenarios, based on the spatial geometric relationship between the beam and the blade, construct the blade straight line equation and the beam straight line equation, solve the intersection point simultaneously to obtain the upper and lower limits of the beam clearance value, and then combine the values ​​calculated in real time in step S3. Substitute the reasonable range of the clearance value, compensate for the lower limit of the range, and set reasonable ranges for the clearance values ​​of the three beams respectively; S6. All data at this moment satisfy the constraints and empirical rules, and output the valid net clearance value; S5 includes the following steps: S5.1 Linear Constraint Model Construction: Based on the included angle of the three beams and the blade's geometric confidentiality parameters, the blade's linear equation and the beam's linear equation are constructed. Solving these equations simultaneously yields the upper and lower limits of the three beams' clearance values. The upper limit constraint formula for the clearance value, where the target beam m hits the blade tip, is as follows: ; in, The vertical height from the blade tip to the radar mounting surface; Let m be the angle between the beam and the vertical direction; The horizontal distance between the radar and the center of the tower. The equivalent radius of the blade tip sweeping across the plane of the tower; When another beam n hits the blade tip, the equation of the blade's straight line is: ； in, Leaf slope , The x-coordinate is the root of the leaf. The vertical coordinate is the base of the leaf. Let x be the x-coordinate of the blade tip point triggered by beam n. Let be the ordinate of the blade tip point triggered by beam n; The equation of the straight line for beam m is: ； By combining the linear equations of the blade and the beam m, the x-coordinate of the intersection point can be obtained. We can obtain: ； Substituting the x-coordinate of the intersection point into the clearance value formula, we can obtain the lower clearance value of beam m as follows: ； in, The x-coordinate of the intersection point of the blade line and the beam line m; When beams m and n fail to hit the blade tip, it is necessary to solve the equations for the blade and the beam simultaneously to calculate the upper limit of the clearance value corresponding to the beam. Lower limit of net air value ; S5.2 Blade Deformation Constraint Superposition: The real-time calculated nonlinear blade deformation variable ΔL is substituted into the rationality interval, and adaptive compensation is performed on the interval. The rationality interval after compensation is: ； S5.3 Verification: At this moment, all three beams fall within their respective reasonable intervals, satisfying the single-beam spatial constraint requirements; In S6, the constraints and empirical rules include spatial association rules, temporal gradient rules, data validity rules, and numerical lower limit rules. In the spatial association rules, if beam 1 is triggered, beams 2 and 3 must also be triggered, and if beam 2 is triggered, beam 3 must also be triggered. In the temporal gradient rules, the scene at the current moment and the scenes at the previous and next moments are adjacent level jumps, and the clearance value is only allowed to jump to the adjacent scene, and cross-scene abrupt changes are prohibited. In the data validity rules, all triggered beam data are valid data, and invalid values ​​are prohibited. In the numerical lower limit rules, the clearance value of a single beam must not be lower than the minimum clearance value when all three sets of beams are triggered.

2. The clearance value optimization method based on three-beam spatial constraints according to claim 1, characterized in that, S2 includes the following steps: S2.1, Preset threshold I based on backlight intensity th Remove low signal-to-noise ratio data, remove invalid ranging values ​​affected by weather, and mark the validity status of the three sets of beam data; S2.

2. Based on radar geometry, obtain the real-time clearance value of beam n (n=1,2,3). The formula is expressed as: ； in, The effective ranging value of beam n Let n be the angle between the beam and the vertical direction. For beam The horizontal component.

3. The clearance value optimization method based on three-beam spatial constraints according to claim 1, characterized in that, In S3, the common rationality constraints for blade triggering include blade ranging range constraints and deformation adaptive correction. Among the blade ranging range constraints, when the beam triggers the blade, the distance L from the target point to the blade root must satisfy the following formula: ； in, The effective length of the blade. , The coordinates of the target point, The coordinates of the leaf root are... The nonlinear deformation of the blades is determined by wind speed. Rotation speed Pitch angle The deformation of the blade tip towards the tower is calculated in real time and used to correct the effective detection range of the blade tip and compensate for the positional offset caused by the bending of the blade. The deformation adaptive correction is used to adaptively adjust the reasonable range of the net clearance value based on the real-time calculated ΔL.

4. The clearance value optimization method based on three-beam spatial constraints according to claim 1, characterized in that, In S4, with beam 3 as the priority core, the triggering scenarios are divided into the first scenario, the second scenario, the third scenario, and the fourth scenario. In the first scenario, beam 3 does not trigger the blades; in the second scenario, only beam 3 triggers the blades; in the third scenario, beams 2 and 3 trigger the blades, while beam 1 does not trigger; in the fourth scenario, beams 1, 2, and 3 all trigger the blades.

5. The clearance value optimization method based on three-beam spatial constraints according to claim 1, characterized in that, In S6, the effective clearance value follows the minimum function principle, selecting the minimum clearance value of the effective trigger beam as the final reliable clearance value. The formula is: ； in, The clearance value for beam 2. This is the clearance value for beam 3.

6. The clearance value optimization method based on three-beam spatial constraints according to claim 1, characterized in that, In S1, the laser air clearance radar emits a detection laser beam toward the wind turbine blades to actively detect the air clearance status of the blades. During the signal propagation and scattering process, the air clearance radar collects data from the received echoes.

7. A clearance value optimization system based on a three-beam spatial constraint-based clearance value optimization method according to any one of claims 1-6, the system comprising a data acquisition module, a data preprocessing module, a constraint determination module, a scene recognition module, a constraint configuration module, and a clearance value calculation module, wherein the data acquisition module is used to monitor and acquire data from three sets of beams, the data preprocessing module is used to preprocess the acquired data and then perform preliminary calculation of the clearance value, the constraint determination module is used to perform verification according to preset constraint conditions, the scene recognition module is used to determine the triggering state of the beams and then identify the triggering scene, the constraint configuration module is used to configure the legal and reasonable value ranges corresponding to the real-time clearance values ​​of the three sets of beams respectively, and the clearance value calculation module is used to output the final clearance value after integrating empirical rules.

Images