Cabin type laser radar wind field detection method and system based on multi-beam expansion

Abstract

The invention discloses a cabin type laser radar wind field detection method and system based on multi-beam expansion. The cabin type laser radar wind field detection method and system are low in cost, high in sampling density, small in fitting error and high in measurement precision. The method comprises the following steps: transmitting a plurality of groups of uniformly distributed radial detection wave beams through a cabin type wind measurement laser radar, and carrying out omnibearing scanning on a target wind field; collecting radial wind speed observation data of each group of beams, performing nonlinear fitting processing on the observation data based on a cosine fitting model, and determining a mapping relation between fitting parameters and azimuth angles; the real wind speed and wind direction of the target wind field are calculated through the fitting model; the system comprises a transmitting module (1), a receiving module (2), a fitting calculation module (3) and an output module (4). The method is applied to the technical field of wind measurement laser radars.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wind measurement lidar, in particular to a cabin type lidar wind field detection method and system based on multi-beam extension, and belongs to the cross technical field of industrial sensing technology service and industrial big data intelligent processing. BACKGROUND

[0002] As a high-precision wind field measurement device, wind measurement lidar has been widely used in wind power generation, aerospace, meteorological observation and other fields due to its fast response speed, wide measurement range and strong anti-interference ability. Among them, the cabin type wind measurement lidar is directly installed on the top of the wind turbine cabin, which can obtain the wind field data in front of the impeller in real time, and provide key support for the variable pitch control and power regulation of the wind turbine, which is of great significance to improve the power generation efficiency and operation safety of the wind turbine.

[0003] The existing cabin type wind measurement lidar generally adopts a 4-beam detection scheme, which transmits 4 groups of radial beams at different azimuth angles, collects radial wind speed data, and then performs wind field inversion. However, this scheme has obvious technical defects: the sampling density of the 4-beam is low, the azimuth angle coverage interval is large (usually 90°), it is difficult to fully capture the spatial distribution characteristics of the wind field, resulting in large fitting error in the wind field inversion process; at the same time, the fitting algorithm used in the traditional 4-beam scheme has poor adaptability to sparse data, and in complex wind field environment, the detection accuracy of wind speed and direction cannot meet the high-precision control requirements. According to the actual application data, the wind speed fitting goodness R² of the existing 4-beam wind measurement lidar is about 0.9830, and the wind direction fitting goodness R² is about 0.9435, which still has room for improvement.

[0004] In order to solve the above problems, there is an urgent need in the field for a wind measurement scheme that can improve the beam sampling density and optimize the fitting algorithm. Based on this, the present application proposes to expand the number of beams from 4 to multiple, combined with an optimized radial wind speed cosine fitting model, to realize high-precision detection of wind field parameters and make up for the shortcomings of the existing technology.

[0005] Under the background of building a new power system and promoting the digital transformation of energy, wind farms as important distributed energy nodes, their operation data are the key input source of industrial artificial intelligence network system. However, the noise and uncertainty of the original wind measurement data seriously restrict the efficiency of the upper industrial information and data processing services (such as intelligent operation and maintenance, power accurate prediction). Therefore, there is an urgent need for an industrial data integration service module with strong robustness that can be embedded in the edge computing or cloud data platform of the wind farm, to realize real-time and efficient cleaning and refining of the original sensing data. SUMMARY

[0006] This invention aims to provide a robust intelligent processing solution for wind farm sensor data in industrial applications. This solution can serve as a core functional module in the industrial artificial intelligence network system of the wind power industry, improving the overall quality of industrial data integration services.

[0007] The purpose of this invention is to provide a low-cost, high-sampling-density, low-fitting-error, and high-precision nacelle-type lidar wind field detection method and system based on multi-beam extension. By increasing the number of radial beams to improve wind field sampling coverage, and combining this with cosine fitting algorithm optimization, high-precision measurement of wind speed and direction is achieved, meeting the high-precision wind measurement requirements of scenarios such as wind turbine generators, while maintaining low cost.

[0008] The technical solution adopted in this invention is a wind field detection method based on a nacelle-mounted lidar with multi-beam spread, which includes the following steps: S1. The transmitting module of the cabin-type wind-measuring lidar generates 4×k groups of radial detection beams, where k is a natural number. The 4×k groups of beams are evenly distributed along the azimuth angle λ, and the azimuth angle coverage range is 0°~360°. S2. Control 4×k groups of probe beams to simultaneously scan the target wind field, and collect radial wind speed observation data of each group of beams through the receiving module. u i ,in i =1,2,...,4×k; S3. Construct a radial wind speed cosine fitting model based on radial wind speed observation data. u i Perform nonlinear fitting and solve for the fitting parameters. m , n and λ max ; S4. Calculate the true wind speed of the target wind field based on the fitted parameters. V and wind direction θ Output wind field detection results.

[0009] Furthermore, in step S3, the expression for the radial wind speed cosine fitting model is: v fit = m + n· cos( λ - λ max ), in, v fit To fit the radial wind speed, λ The beam azimuth angle, m For the fitting constant term, n To fit the amplitude term,λ max To fit the peak value of the azimuth angle.

[0010] Furthermore, in step S1, the transmission power of the 4×k groups of detection beams is consistent, the beam divergence angle is set to 0.5°~1.0°, and the detection distance range is 10m~200m.

[0011] Furthermore, in step S3, the least squares method is used to solve for the fitting parameters. m , n and λ max The objective function is: min∑( u i -( m + n· cos( λ i - λ max )))², i=1 to 4×k, in, u i For the first i Radial wind speed observed by beamforming λ i For the first i Azimuth angle of the group beam.

[0012] Furthermore, in step S4, the actual wind speed V The calculation expression is: , The expression for calculating the true wind direction θ is: when m When >0, θ = λ max +90° when m When <0, θ= λ max -90°.

[0013] Furthermore, in step S2, the radial wind speed observation data is collected. u i The signal-to-noise ratio must be ≥30dB, and the data sampling frequency must be 10Hz~50Hz.

[0014] Furthermore, the method also includes step S5: fitting the true wind speed V and wind direction θ Smoothing is performed, and a moving average algorithm is used to eliminate random errors. The sliding window size is set to 5 to 10 sampling points.

[0015] A nacelle-type wind-measuring lidar system for implementing the multi-beam spread-based nacelle-type lidar wind field detection method as described above, the system comprising a transmitting module, a receiving module, a fitting calculation module, and an output module; The transmitting module is used to generate 4×k groups of uniformly distributed radial probe beams; The receiving module is used to collect radial wind speed observation data for each group of beams; The fitting calculation module has a built-in radial wind speed cosine fitting model, which is used to solve the fitting parameters and calculate the actual wind speed and wind direction; The output module is used to output the wind field detection results.

[0016] Furthermore, the emission module includes a laser emission unit, which is mounted on a turntable driven by a motor. The motor drives the turntable to rotate and causes the laser emission unit to rotate, thereby achieving 4×k sets of laser beam detection and scanning. The wavelength of the laser emission unit is set to 1550nm or 1064nm.

[0017] Furthermore, the fitting calculation module also includes a data verification unit, used to verify the validity of the observed data and remove abnormal data that exceed a preset threshold, wherein the preset threshold is ±3 times the standard deviation.

[0018] This invention, through the coordinated design of multi-beam spread and optimized fitting algorithms, has the following significant advantages compared to existing four-beam wind measurement schemes: (1) Improve wind field sampling density and coverage integrity: The number of beams is expanded from 4 groups to 4×k groups, and the azimuth interval is reduced from 90° to a small angle, which significantly improves the spatial sampling density of the wind field. More sampling points can more comprehensively reflect the spatial distribution characteristics of the wind field, reduce the fitting deviation caused by sparse sampling, and provide a data basis for high-precision wind field inversion.

[0019] (2) Optimize fitting accuracy and measurement reliability: The radial wind speed cosine fitting model is adopted. This model can accurately describe the variation law of the radial component of the wind field with the azimuth angle. Combined with the least squares method to solve the fitting parameters, the goodness of fit is significantly improved. Compared with the 4-beam scheme, the observation data of 4×k beams provides more constraints for fitting, reduces the fitting error, and makes the goodness of fit of wind speed R² reach more than 0.98 and the goodness of fit of wind direction R² reach more than 0.94. The measurement results are closer to the real wind field.

[0020] (3) Enhanced adaptability to complex wind fields: The multi-beam all-round coverage design of this invention can effectively capture the dynamic changes of complex wind fields (such as turbulence and wind shear). The fitting model is more adaptable to multi-sampling data and can maintain high measurement accuracy even in non-uniform wind field environments. However, the existing 4-beam scheme is prone to fitting distortion in complex wind fields due to insufficient sampling points, resulting in poor measurement stability.

[0021] (4) Practicality for nacelle-type scenarios: The beam parameters (detection distance 10m~200m, divergence angle 0.5°~1.0°) are specifically designed for nacelle-type installation scenarios. The synchronous scanning of multiple beams does not affect the real-time measurement (sampling frequency 10Hz~50Hz). Furthermore, by setting up a combination of motor and turntable, there is no need to increase the number of laser emission units. While maintaining low cost, it can meet the real-time requirements of wind turbine pitch control for wind field data. At the same time, it does not require a significant increase in equipment size and energy consumption, and has good engineering feasibility. Attached Figure Description

[0022] Figure 1 This is a simplified structural block diagram of the system of the present invention; Figure 2 This is a simplified flowchart of the method of the present invention; Figure 3 This is a diagram of the laser beam model in this invention; Figure 4 It is a fitted curve of wind speed versus azimuth angle; Figure 5 The images show the wind speed and wind direction fitting curves obtained using existing technology with four sets of beams. The left image shows the wind speed fitting result (wind direction 100°~150°), and the right image shows the wind direction fitting result (100°~150°). Figure 6 The wind speed and wind direction fitting result curves obtained by the present invention using 12 groups of beams are shown. The left figure is the wind speed fitting result (wind direction 100°~150°), and the right figure is the wind direction fitting result (100°~150°). Detailed Implementation

[0023] like Figures 1-6 As shown in this embodiment, the number of beams emitted by the 12 groups of beams as laser emitting units will be used to illustrate the present invention in more detail. The nacelle-type wind-measuring lidar system used in this invention includes a transmitting module 1, a receiving module 2, a fitting calculation module 3, and an output module 4. The transmitting module 1 is used to generate 12 sets of uniformly distributed radial probe beams; The receiving module 2 is used to collect radial wind speed observation data for each group of beams; The fitting calculation module 3 has a built-in radial wind speed cosine fitting model, which is used to solve the fitting parameters and calculate the actual wind speed and wind direction. The output module 4 is used to output the wind field detection results.

[0024] Specifically, the emission module 1 includes a laser emission unit mounted on a motor-driven turntable. The motor drives the turntable to rotate, which in turn rotates the laser emission unit, enabling 12 sets of laser beam detection scanning. The wavelength of the laser emission unit is set to 1550nm or 1064nm. The fitting calculation module 3 also includes a data verification unit for verifying the validity of the observation data and removing abnormal data exceeding a preset threshold, where the preset threshold is ±3 standard deviations.

[0025] The method for wind field detection using a cabin-mounted lidar system with multi-beam extension, as described above, includes the following steps: S1. The transmitting module of the cabin-type wind-measuring lidar generates 4×k groups of radial detection beams, where k is a natural number. The 12 groups of beams are evenly distributed along the azimuth angle λ, with an azimuth angle coverage range of 0° to 360° and an azimuth angle interval of 30° between adjacent beams. S2. Control 12 sets of detection beams to simultaneously scan the target wind field, and collect radial wind speed observation data of each set of beams through the receiving module. u i ,in i =1,2,...,12; S3. Construct a radial wind speed cosine fitting model based on radial wind speed observation data. u i Perform nonlinear fitting and solve for the fitting parameters. m , n and λ max ; S4. Calculate the true wind speed of the target wind field based on the fitted parameters. V and wind direction θ Output wind field detection results.

[0026] In step S3, the expression for the radial wind speed cosine fitting model is: v fit = m + n· cos( λ - λ max ), in, v fit To fit the radial wind speed, λ The beam azimuth angle, m For the fitting constant term,n To fit the amplitude term, λ max To fit the peak value of the azimuth angle.

[0027] In step S1, the 12 sets of detection beams have the same transmission power, the beam divergence angle is set to 0.5° to 1.0°, and the detection distance ranges from 10m to 200m.

[0028] In step S3, the least squares method is used to solve for the fitting parameters. m , n and λ max The objective function is: min∑( u i -( m + n· cos( λ i - λ max )))², i=1~12, in, u i For the first i Radial wind speed observed by beamforming λ i For the first i Azimuth angle of the group beam.

[0029] In step S4, the actual wind speed V The calculation expression is: , The expression for calculating the true wind direction θ is: when m When >0, θ = λ max +90° when m When <0, θ= λ max -90°.

[0030] In step S2, the radial wind speed observation data is collected. u i The signal-to-noise ratio must be ≥30dB, and the data sampling frequency must be 10Hz~50Hz.

[0031] The method also includes step S5: fitting the true wind speed V and wind direction θSmoothing is performed using a moving average algorithm to eliminate random errors, with the moving window size set to 5-10 sampling points. The moving average algorithm reduces random fluctuations in a single sampling by weighting the wind field data (actual wind speed V, wind direction θ) from multiple consecutive sampling points, thereby improving data stability. The specific steps are as follows: Data sequence determination: Using the continuous wind field data calculated in step S4 as the processing object, an original data sequence is constructed. Let the sampling frequency be f (unit: Hz, ranging from 10Hz to 50Hz, e.g., f=20Hz in Example 1 and f=30Hz in Example 2). Then, one set of data (containing one wind speed value) is generated every 1 / f seconds. V j and 1 wind direction value θj ), forming the original sequence: Original wind speed sequence: V =[ V 1, V 2, V 3,..., V j ,..., V n ],in j For data sequence number ( j =1,2,..., n ), n Total data volume; Original wind direction sequence: θ=[ θ 1 , θ 2 , θ 3 ,..., θ j ,..., θ n ].

[0032] Sliding window parameter settings: Based on the dynamic characteristics of the wind field and the sampling frequency, set the sliding window size to N (unit: sampling points, value 5-10, such as N=8 in Example 1). The window must meet the principle of "no crossing of space and no repeated calculation of edges": that is, the window starts from the first data point, and moves one sampling point along the data sequence direction each time, until the end of the window covers the last data point. V n or θ n ).

[0033] Data extraction and validity verification within the window: For each sliding window, the corresponding data segment is extracted, and outlier screening is performed using the "data verification unit" in the fitting calculation module. Extracting window data segments: For the k-th window (k=1,2,...,n-N+1), extract the wind speed segment. V k ~ V k+N-1 Wind direction section θ k ~ θ k+N-1 .

[0034] Outlier removal: Calculate the mean (μ) and standard deviation (σ) of the data segment in the window. If a data point exceeds the range of [μ-3σ, μ+3σ], it is determined to be an outlier and replaced with the mean of other valid data in the window to ensure that the number of data points in the window is always N.

[0035] Moving average calculation: The "equal-weighted moving average" method is used (to adapt to the real-time requirements of the cabin-style scenario and avoid delays caused by complex weighted calculations). The average values ​​of wind speed and wind direction data within the window are calculated separately to obtain the smoothed data for that window. Smoothed wind speed V k' : V k' =( V k + V k+1 +...+ V k+N-1 ) / N Smoothed wind direction θ k' : θ k' =( θ k + θ k+1 +...+ θ k+N-1 ) / N Special Note: If the wind direction data jumps across 0° / 360° (e.g., ... θ k =359° θ k+1 =2°), all wind direction values ​​need to be converted to the "-180°~180°" coordinate system first (360° corresponds to 0°, 359° corresponds to -1°) before calculating the average, and finally the result is converted back to the "0°~360°" coordinate system to avoid angle calculation errors.

[0036] Smoothed sequence output: Repeat steps 3-4 until all sliding windows have been calculated, resulting in a smoothed data sequence with the same length as the original data sequence. Smoothed wind speed sequence: V' =[ V1' , V 2' ,..., V n-N+1' ,..., V n' ]; Smoothed wind direction sequence: θ' =[ θ 1' , θ 2' ,..., θ n-N+1' ,..., θ n' ].

[0037] This sequence is the final wind field data after eliminating random errors, and it is used in the input / output module for scenarios such as wind turbine pitch control and power regulation.

[0038] For example, in Example 1, a sliding window with N=8 is used for 10 sets of raw wind speed data: The values ​​[10.05, 10.03, 9.98, 10.01, 10.02, 9.99, 10.04, 10.00, 10.02, 10.01] are processed. The first window calculates the mean (10.05+10.03+9.98+10.01+10.02+9.99+10.04+10.00) / 8≈10.01, and the second window calculates the mean (10.03+9.98+10.01+10.02+9.99+10.04+10.00+10.02) / 8≈10.01. The final output is a smoothed wind speed sequence, ensuring that the error is stable within ±0.01m / s, which meets the requirements of high-precision wind measurement.

[0039] The above smoothed wind speed sequence V' and wind direction sequence θ' This constitutes a high-quality industrial field sensor data stream. This data stream can be directly connected to the wind farm's central monitoring system or edge computing gateway, serving as a key input parameter to drive various industrial artificial intelligence applications. For example, in wind turbine intelligent control systems, this data provides accurate basis for real-time pitch and yaw decisions; in wind farm-level power prediction models, this data, as a core feature input, can significantly improve the accuracy of short-term and ultra-short-term power predictions; in equipment health monitoring systems, long-term stable high-quality wind data helps to more accurately assess wind turbine loads and fatigue damage. Therefore, this invention is not only a data processing algorithm, but also a key data bridge connecting underlying physical sensing and upper-level industrial intelligent services, effectively supporting the construction of an industrial data integration service system in the wind power field.

[0040] The present invention will now be further described with reference to more specific embodiments.

[0041] Example 1:

[0042] This embodiment is applied to the nacelle-type wind measurement scenario of a 1.5MW wind turbine generator set. The actual wind speed of the target wind field is 10m / s, the actual wind direction is 120°, and the environment is a stable wind field (turbulence intensity <0.1).

[0043] Beam parameter settings: The transmitting module has 12 laser emitting units with a wavelength of 1550nm, a transmitting power of 8W, a beam divergence angle of 0.8°, and a detection range of 50m to 150m. The azimuth angles of the 12 beams are... λ i The angles are 0°, 30°, 60°, 90°, 120°, 150°, 180°, 210°, 240°, 270°, 300°, and 330°, respectively, and are evenly distributed between 0° and 360°.

[0044] Data collection: The sampling frequency is set to 20Hz, and 12 beams are sampled (beam number...). i Radial wind speed observation data u i (Unit: m / s), some observation data are shown in the table below: Beam number 1（0°） 2（30°） 3（60°） 4（90°） 5（120°） 6（150°） 7（180°） 8（210°） 9（240°） 10（270°） 11（300°） 12（330°） Observed wind speed 2.5 4.8 7.2 9.5 10.0 9.3 7.0 4.5 2.2 0.5 0.8 1.9 Solving for fitting parameters: Radial wind speed observation data u i and the corresponding azimuth angle λ i Substituting the parameters into the following objective function formula, the least squares method is used to solve for the fitting parameters: min∑( u i -( m + n· cos( λ i - λ max )))², Calculations show that: m =5.02, n =4.98, λ max =120.5°, The fitted model expression is: v fit =5.02+4.98 · cos(λ-120.5°).

[0045] Wind field parameter calculation: Substituting into the formula for calculating the actual wind speed V, we get the actual wind speed as V≈10.03m / s; because m =5.02>0, true wind direction θ for: θ = 120.5° + 90° = 210.5°. Using the wind field coordinate system, the true wind direction is obtained after correction. θ The value is 120.3°, which deviates from the actual wind direction of 120° by 0.3°.

[0046] Data smoothing: A sliding average algorithm with a sliding window size of 8 sampling points was used to smooth 10 sets of continuously calculated wind speed and wind direction data. The final output results were: wind speed 10.01 m / s and wind direction 120.2°.

[0047] Accuracy verification: The wind speed error measured in this study was |10.01-10|=0.01m / s, the wind direction error was |120.2-120|=0.2°, the wind speed goodness of fit R²=0.986, and the wind direction goodness of fit R²=0.945. Compared with the existing four-beam scheme, the error was significantly reduced and the accuracy was significantly improved.

[0048] The high-precision, low-latency data processing capabilities verified in this embodiment fully demonstrate the applicability and reliability of the present invention in complex industrial environments. The processed data can directly serve the digital and intelligent operation of wind farms, making it an indispensable part of building an efficient and reliable industrial artificial intelligence network system.

[0049] Example 2:

[0050] This embodiment is applied to the nacelle-type wind measurement scenario of a 2.5MW wind turbine generator set. The target wind field is a complex turbulent wind field (turbulence intensity 0.15), the actual wind speed is 8m / s, and the actual wind direction is 90°.

[0051] Beam parameters and data acquisition: The beam parameters are the same as in Example 1, the sampling frequency is 30Hz, and 12 sets of observed wind speeds are collected. u i as follows: Beam number 1（0°） 2（30°） 3（60°） 4（90°） 5（120°） 6（150°） 7（180°） 8（210°） 9（240°） 10（270°） 11（300°） 12（330°） Observed wind speed 1.8 4.2 6.5 7.9 7.5 6.1 4.0 1.7 0.3 1.2 2.5 3.8 Fitting and Calculation: The fitting parameters are obtained by solving: m =3.98, n =3.95, λ max =91.2°, The actual wind speed V≈8.02m / s The wind direction θ = 91.2° + 90° = 181.2°. After correction using the wind field coordinate system, the true wind direction is obtained. θ The value is 89.8°, which deviates from the actual wind direction of 90° by 0.2°.

[0052] Result verification: The final smoothed wind speed was 8.01 m / s, the wind direction was 89.9°, the wind speed error was 0.01 m / s, the wind direction error was 0.1°, the wind speed goodness of fit R²=0.984, and the wind direction goodness of fit R²=0.944. It still maintains high measurement accuracy in complex turbulent wind fields, which verifies the effectiveness of the invention.

[0053] Depend on Figure 5 and Figure 6 As can be seen from the comparison, the present invention increases the number of radial beams, which can effectively improve the accuracy of wind field detection.

[0054] Finally, it should be emphasized that the above description is only a preferred embodiment of the present invention and 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 method for wind field detection using a cabin-mounted lidar based on multi-beam spread, characterized in that, The method includes the following steps: S1. The transmitting module of the cabin-type wind-measuring lidar generates 4×k groups of radial detection beams, where k is a natural number. The 4×k groups of beams are evenly distributed along the azimuth angle λ, and the azimuth angle coverage range is 0°~360°. S2. Control 4×k groups of probe beams to simultaneously scan the target wind field, and collect radial wind speed observation data of each group of beams through the receiving module. u i ,in i =1,2,...,4×k; S3. Construct a radial wind speed cosine fitting model based on radial wind speed observation data. u i Perform nonlinear fitting and solve for the fitting parameters. m , n and λ max ; S4. Calculate the true wind speed of the target wind field based on the fitted parameters. V and wind direction θ Output wind field detection results.

2. The method for wind field detection using a nacelle-type lidar based on multi-beam spread as described in claim 1, characterized in that, In step S3, the expression for the radial wind speed cosine fitting model is: v fit = m + n· cos( λ - λ max ), in, v fit To fit the radial wind speed, λ The beam azimuth angle, m For the fitting constant term, n To fit the amplitude term, λ max To fit the peak value of the azimuth angle.

3. The method for wind field detection using a nacelle-type lidar based on multi-beam spread as described in claim 1, characterized in that, In step S1, the transmission power of the 4×k groups of detection beams is the same, the beam divergence angle is set to 0.5°~1.0°, and the detection distance range is 10m~200m.

4. The method for wind field detection using a nacelle-type lidar based on multi-beam spread as described in claim 1, characterized in that, In step S3, the least squares method is used to solve for the fitting parameters. m , n and λ max The objective function is: min∑( u i -( m + n· cos( λ i - λ max )))², i=1 to 4×k, in, u i For the first i Radial wind speed observed by beamforming λ i For the first i Azimuth angle of the group beam.

5. The method for wind field detection using a nacelle-type lidar based on multi-beam spread as described in claim 1, characterized in that, In step S4, the actual wind speed V The calculation expression is: ， The expression for calculating the true wind direction θ is: when m When >0, θ = λ max +90° when m When <0, θ= λ max -90°.

6. The method for wind field detection using a nacelle-type lidar based on multi-beam spread as described in claim 1, characterized in that, In step S2, the radial wind speed observation data is collected. u i The signal-to-noise ratio must be ≥30dB, and the data sampling frequency must be 10Hz~50Hz.

7. The method for wind field detection using a nacelle-type lidar based on multi-beam spread as described in claim 1, characterized in that, The method also includes step S5: fitting the true wind speed V and wind direction θ Smoothing is performed, and a moving average algorithm is used to eliminate random errors. The sliding window size is set to 5 to 10 sampling points.

8. A nacelle-type wind-measuring lidar system for implementing the nacelle-type lidar wind field detection method based on multi-beam spread as described in any one of claims 1 to 7, characterized in that, The system includes a transmitting module (1), a receiving module (2), a fitting calculation module (3), and an output module (4). The transmitting module (1) is used to generate 4×k groups of uniformly distributed radial probe beams; The receiving module (2) is used to collect radial wind speed observation data of each group of beams; The fitting calculation module (3) has the radial wind speed cosine fitting model built in, which is used to solve the fitting parameters and calculate the real wind speed and wind direction; The output module (4) is used to output the wind field detection results.

9. The cabin-type wind-measuring lidar system according to claim 8, characterized in that, The emission module (1) includes a laser emission unit, which is set on a turntable driven by a motor. The motor drives the turntable to rotate and drives the laser emission unit to rotate, thereby realizing 4×k groups of laser beam detection and scanning. The wavelength of the laser emission unit is set to 1550nm or 1064nm.

10. The cabin-type wind-measuring lidar system according to claim 8, characterized in that, The fitting calculation module (3) also includes a data verification unit, which is used to verify the validity of the observation data and remove abnormal data that exceed a preset threshold, wherein the preset threshold is ±3 times the standard deviation.

Images