487 results about "Curve fitting" patented technology

The invention relates to the technical field of metering pumps, in particular to an external meshing gear metering pump flow calculation method based on linear regression and curve fitting, which comprises the following steps: generating an integrated data set by monitoring the flow rate, pressure, temperature and viscosity in real time, setting flow rate, pressure and temperature thresholds based on the data set, judging the flow state and recording abnormity, and calculating the flow rate of the external meshing gear metering pump. Flow velocity and viscosity are selected for flow calculation, a corrected flow value is generated, whether flow is normal or not is analyzed and determined, a flow stability evaluation result is generated, and a visual interface is created to display parameters and a calculation result. According to the method, the multiple sensors are installed at the inlet and the outlet of the metering pump, the flow speed, the pressure, the temperature and the viscosity are monitored in real time, an integrated data set provides a precise basis for flow calculation, a threshold value is set, data are monitored, it is ensured that parameters exceeding the normal range can be recorded and analyzed in time, and the process improves the accuracy of flow calculation and improves the working efficiency. And flow evaluation is optimized, and the operation safety and efficiency are improved.

The invention provides an adaptive welding seam detection and three-dimensional reconstruction method based on deep learning and binocular vision. The adaptive welding seam detection and three-dimensional reconstruction method comprises the steps of S1, collecting samples and making a training data set; s2, the picture of the sample to be welded is processed, a feature region is recognized, the image quality of the region to be welded is analyzed and evaluated through wavelet transform and local variance, and the noise level and the contrast ratio are calculated; s3, dynamically generating edge detection parameters and model fitting parameters according to the image quality; s4, using an edge detection algorithm to extract edge point cloud of the welding seam area; s5, performing RANSAC linear fitting, weighted least square fitting and polynomial curve fitting on the edge point cloud in parallel; s6, selecting an optimal fitting result based on an image quality adaptive dynamic scoring model; and S7, carrying out three-dimensional coordinate conversion in combination with the three-dimensional matching model IGEV-Stereo, and outputting a final welding seam three-dimensional coordinate. According to the invention, automatic detection of the position and size of the welding seam can be efficiently and accurately realized.

The invention discloses a machine vision-based wood surface defect online detection method and system. The method comprises the following steps of: acquiring surface image data of wood, performing noise reduction on the data by utilizing adaptive median filtering, eliminating surface texture interference based on curve fitting, and performing multi-scale feature extraction on initial wood surface image data by utilizing an improved ResNet-18 network; and through weighted fusion of multispectral channels, extracting difference features of the wood surface, calculating an adaptive threshold segmentation defect area based on a gray average and a standard deviation of initial wood surface image data, performing optimization by using a Canny operator and morphological operation, and identifying segmented image data and a three-dimensional feature matrix by using a YOLOv8s model. And classifying and positioning the defects. According to the method, rapid and accurate classification and positioning of various wood defects are achieved, and compared with a traditional detection method, the detection efficiency and the detection accuracy are improved.

The invention discloses an unmanned aerial vehicle transformer substation inspection path planning method and device, equipment and a medium. The method comprises the following steps: firstly, generating a grid map; dividing the substation inspection space into a plurality of sub-regions; establishing an electromagnetic exposure intensity model, and setting an electromagnetic exposure intensity safety threshold; randomly sampling a plurality of nodes in each divided sub-region, constructing a random geometric graph, and generating an initial inspection path by adopting an RRT algorithm; optimizing the generated initial inspection path by using a path pruning optimization algorithm and a gradient descent convex optimization algorithm, and smoothing the path into a continuous curve by using a B-spline curve fitting technology to obtain a final optimized inspection path; according to the invention, through an overall strategy of first region division and then path planning, random geometric graph tree building, path pruning optimization, gradient curve smoothing and key region guiding sampling mechanisms are fused, and the efficient, stable and high-environment-adaptability unmanned aerial vehicle three-dimensional path planning method for the substation complex scene is realized.

The invention relates to the technical field of bridge monitoring, and discloses a real-time bridge monitoring and evaluating method based on beam end displacement and rotation angle, which comprises the following steps: acquiring beam end displacement and rotation angle data acquired by a displacement sensor and a rotation angle sensor in a continuous monitoring period, and filtering and normalizing the acquired data to obtain processed data; constructing a discrete curve and performing polynomial curve fitting by using a settlement trajectory multi-cycle residual error clustering identification method, calculating a single-cycle residual error sequence, determining an edge fluctuation trend point by comparing adjacent cycle residual error values, and performing anomaly marking; inputting the marked abnormal data into an abnormal recognition model, predicting the risk level of the bridge structure through the abnormal recognition model, and outputting an abnormal recognition result; outputting a comprehensive structure response diagnosis report according to the abnormity identification result; the precision of monitoring and evaluation is improved, and safe operation of the bridge is guaranteed.

The invention relates to the technical field of hydrological data analysis, in particular to a similar flood-based water level flow relationship intelligent extension method. The method comprises the following steps: acquiring actually measured water level and flow data and performing curve fitting; carrying out three-property test on the fitting curve, and if the fitting curve does not meet the requirement, adopting multi-point sectional type fitting to obtain an output curve; screening a similar curve based on a historical flood graphic library, and carrying out extension processing to form an extended water level flow relation curve; and the real-time water level is subjected to flow pushing by utilizing the extension curve, so that accurate prediction and dynamic extension of the water level and flow relationship are realized. According to the method, the reliability of the curve is guaranteed through actually measured water level flow data fitting and segmented optimization in combination with three-property test, and intelligent complementation and real-time correction under the condition of insufficient data are realized by utilizing a historical flood similar curve extension and dynamic matching mechanism, so that the precision, continuity and adaptability of the water level flow relation curve are remarkably improved; and the accuracy and practicability of the hydrological plug flow process are ensured.

The invention discloses an accurate monitoring system for linear deformation of a loess tunnel portal side upward slope, and relates to the technical field of geotechnical engineering and surveying and mapping. The system comprises a sensing network layer, a data acquisition and transmission layer, a cloud platform and data analysis layer and an application display layer, displacement data of discrete monitoring points are reconstructed into continuous full-field displacement and strain distribution through Kriging interpolation, then a displacement vector field is regarded as a whole to be analyzed, the maximum shear strain field of the displacement vector field is calculated, an abnormal area with the deformation gradient abrupt change is accurately positioned, and the deformation gradient abrupt change accuracy is improved. The abnormal point sets are merged into independent clusters according to spatial proximity and deformation consistency, and the geometric trajectory of a potential slip crack surface or crack is accurately outlined through curve fitting, so that the technical path fundamentally changes the limitation of point-mode monitoring, and can automatically identify a hidden linear failure trend which is difficult to find by naked eyes in a non-contact manner, so as to improve the detection accuracy of the hidden linear failure trend. And refined data support is provided for slope stability evaluation.

The invention discloses a containment failure probability curve fitting method and system based on a random forest, and relates to the technical field of nuclear power safety. Comprising the following steps: acquiring key parameters influencing ultimate bearing capacity of a containment, determining a distribution interval of the key parameters, and establishing a key parameter table; performing finite element calculation on the basis of the key parameter table, obtaining ultimate bearing capacity of the containment vessel under different parameter combinations and temperature loads, and establishing a database; performing normalization preprocessing on the database to form a training data set; establishing a random forest model and performing regression model training by using the training data set; inputting actual design parameters of the containment and the temperature required to be evaluated into the trained model to obtain an ultimate bearing capacity predicted value of the containment; and fitting of a failure probability curve is carried out based on the central limit theorem. According to the method, complex probabilistic safety analysis is converted into a standardized calculation task, and the implementation threshold of nuclear power plant design verification and safety review is greatly reduced.

The invention relates to the technical field of path planning, and discloses a multi-sensor dynamic path planning method and system for a teaching-free welding robot of a large iron tower component, and the method comprises the steps: collecting multi-sensor synchronous data through a laser sensor, a visual sensor and an arc sensor, and solving segmented calibration parameters; calculating three-dimensional position data of the welding seam by combining the segmented calibration parameters and the synchronous data of the multiple sensors; performing error compensation on the three-dimensional position data of the welding seam at the segmented switching point to obtain a welding seam track position; according to the method, error prediction and real-time correction are carried out at the segmented switching points, accumulated positioning deviation is effectively restrained, high-precision teaching-free welding of the long-distance welding seam of the large iron tower component is achieved, and the welding quality of the long-distance welding seam of the large iron tower component is improved. The technical problems that a traditional method is low in precision and poor in efficiency are solved.

According to the method for designing the reasonable arch axis of the large-span arch bridge based on the hybrid calculation algorithm and the self-adaptive load distribution, an analytic equation method, a curve fitting method and an intelligent optimization algorithm are combined, and by selecting the initial arch axis, introducing a dynamic load distribution model and applying a genetic algorithm and a particle swarm optimization algorithm, the reasonable arch axis of the large-span arch bridge is designed. Global and local optimization is carried out, and the optimal shape and load distribution of the arch axis are ensured. And finally outputting design parameters, such as a control point, a bending moment, thrust and the like. According to the method, the design precision, flexibility and safety can be effectively improved, the material utilization rate is optimized, the cost is reduced, and the method is particularly suitable for design of large-span bridges and has high practical value.

The invention relates to the technical field of ocean remote sensing monitoring, in particular to a remote sensing image-based warm discharge water reference temperature self-adaptive extraction method, which specifically comprises the steps of S01, data acquisition and preprocessing, S02, cumulative histogram generation, S03, curve fitting and smoothing, S04, second derivative calculation and S05, reference temperature determination. According to the method, the reference temperature is determined by analyzing the statistical distribution characteristics of the sea surface temperature data field, namely the form of the accumulative histogram, the physical demarcation point of the background water body and the warm drainage area on the temperature distribution is automatically positioned by using the second derivative, the process is completely driven by data, the interference of subjective factors is eliminated, and the accuracy of temperature distribution is improved. The method can automatically adapt to water areas with different geographical forms and different warm drainage diffusion modes, and objectivity and universality of reference temperature extraction are achieved.

The invention discloses a welding seam track identification method for impeller wrap angle welding based on point cloud data, and belongs to the field of machine welding seam identification, and the method comprises the following steps: carrying out filtering, down-sampling and clustering on original scanning point cloud; classifying the point cloud data through a region growing clustering algorithm; determining an optimal interception plane based on principal component analysis, and performing curve fitting on section data to obtain a preliminary main welding seam track; the main welding seam track is projected to a local approximate plane in a segmented mode, and accurate extraction of the main welding seam track is completed; and performing plane fitting and thickness calculation on the wrap angle surface, solving an intersection line in combination with an adjacent local approximate plane, calculating an intersection point of the intersection line and the main welding seam track, and finally generating a wrap angle welding seam track in combination with the direction of the intersection line and the wrap angle thickness. Through cooperative calculation of local plane projection and spatial intersection line solving, the whole process operation from recognition, planning to welding can be automatically completed, and unmanned intelligent welding in the true sense is achieved.

The embodiment of the invention provides a method for determining initial parameters of a transient absorption spectrum kinetics curve, and the method comprises the steps: obtaining an initial absorption point and an absorption attenuation termination point of the transient absorption spectrum kinetics curve, and determining a curve segment between the initial absorption point and the absorption attenuation termination point as a total absorption attenuation interval; carrying out weighting processing on a first-order derivative and a second-order derivative of the transient absorption spectrum dynamic curve to obtain a weighted curvature value sequence; the first N-1 weighted curvature values are selected from the weighted curvature value sequence from large to small, and the total absorption attenuation interval is divided into N absorption attenuation intervals corresponding to curve fitting orders according to sampling points corresponding to the first N-1 weighted curvature values; and determining an initial parameter based on a quantitative relationship of 1 / e of the absorption intensity attenuated to the corresponding interval starting point in each absorption attenuation interval in the exponential fitting model, wherein the quantitative relationship is that the absorption intensity at the interval starting point in each absorption attenuation interval is attenuated to the absorption intensity at the corresponding interval starting point. And the fitting efficiency and accuracy of the dynamic curve are improved.

The invention relates to a dynamic path prediction method for satellite antenna tracking and computer equipment, and the method comprises the steps: collecting satellite antenna tracking angle data which comprise original data and actual data; calculating a first deviation sequence and a second deviation sequence based on the original data and the actual data; performing curve fitting on the first deviation sequence and the second deviation sequence through a fitting algorithm, substituting a fitting result into a pre-constructed mathematical model to obtain a prediction model, and extrapolating and predicting a prediction deviation of a preset time point through the prediction model; and compensating the predicted deviation to original data corresponding to the preset time point, and generating a dynamic range instruction, so that the satellite antenna is aligned with the communication satellite under the action of the dynamic range instruction. And the satellite antenna can continuously align to the real position of the communication satellite.

The invention discloses an AI-based press-fitting curve intelligent analysis and standard curve generation method and system, and relates to the field of press-fitting curve analysis, and the method comprises the steps: constructing an AI classification model and a training set; training an AI classification model by using the training set to obtain a trained AI classification model; performing quality classification on the training set by using the trained AI classification model to obtain a quality classification result; performing standard curve fitting on the press-fitting curve samples with normal quality classification results to obtain a standard curve; and performing quality classification on the to-be-analyzed press fitting curve by using the trained AI classification model to obtain a quality classification result of the to-be-analyzed press fitting curve. The AI intelligent classification diagnosis of the press-fitting curve to be analyzed is automatically realized, manual intervention is reduced, the judgment consistency is improved, and the classification accuracy is improved.

The invention relates to the technical field of fully mechanized coal mining face automation control, and discloses a working face end straightness detection method and device based on a laser radar, and the method comprises the steps: building geometric and radiation references through collecting fixed reference target data; in real-time detection, a displacement vector and an atmospheric attenuation coefficient are calculated by using an instantaneous state of a target, and double-domain dynamic correction of coordinate rigid inverse transformation and intensity threshold adaptive adjustment is performed on an original point cloud. Then, clustering segmentation and slice centroid extraction are carried out on the corrected point cloud, and an ordered feature point set is obtained; and performing multi-factor weighted curve fitting on the feature point set in combination with geometric stability and environmental credibility, constructing a form fitting curve, and outputting a straightness deviation quantitative index according to a comparison result of the curve and an ideal reference straight line. According to the invention, the interference of equipment vibration and dust attenuation on measurement is eliminated, and the robustness and precision of the straightness detection of the scraper conveyor are improved.

The invention discloses an unmanned target aircraft flight path planning method and system. The method comprises the steps of obtaining dynamic battlefield environment data, generating an initial path node sequence by combining an A * algorithm with dynamic battlefield threat data, calculating a communication quality weight sequence by using a path loss model and a Monte Carlo method, and optimizing a path. The method comprises the following steps of: acquiring a path instruction sequence, fusing navigation data through Kalman filtering to evaluate deviation, correcting the path through quadratic programming, generating a smooth path by adopting B-spline curve fitting, evaluating the path execution probability through a Monte Carlo method, and finally generating a path instruction sequence containing speed and direction control signals. According to the method, intelligent path planning in a complex battlefield environment is realized, and the safety, the reliability and the execution efficiency of the path are improved.

The invention discloses a photoelectric angle sensor fuzzy calibration method based on ANSYS simulation, and relates to the technical field of photoelectric angle sensor fuzzy calibration, and the method comprises the following steps: calling node displacement change data along with an angle in sensor simulation output based on an identification result, constructing a periodic displacement offset fitting function family, and carrying out the calibration of the periodic displacement offset fitting function family; performing curve fitting processing on one circle of rotation data so as to determine angle response deflection caused by structure unbalance loading under the condition that a non-axisymmetric loading condition is not set in the simulation process; and based on the determined angle response deflection result, constructing a residual attention enhanced Transform structure, extracting an error change trajectory feature set in a structure deflection envelope, and outputting a drift discrimination result reflecting the structural error drift direction, level and stability. According to the method, the problem of error modeling distortion of the photoelectric angle sensor under the non-axisymmetric load is solved, and error drift feature recognition and precision dynamic regulation and control are realized.

The invention belongs to the field of medical treatment, and relates to a dynamic birth process progress automatic monitoring method, which comprises the following steps: S1, continuously detecting and screening image frames containing intact pubis symphysis and fetal head skull halo; s2, performing standard section screening on the continuous ultrasonic image frames by using a USPNet model based on a Swin Transform architecture; s3, automatically marking a birth process progress angle and a fetal head exposure position through an AI segmentation model, and calculating a fetal head exposure value; s4, driving the three-dimensional fetal head and pelvis model based on the real-time progress angle and fetal head exposure position data, and dynamically rendering and updating a relation graph of the three-dimensional fetal head and pelvis; s5, dynamically displaying the three-dimensional relational graph of the fetal head and the pelvis; and S6, establishing a fetal head descending and uterine contraction signal time sequence curve, and estimating the descending amplitude of the fetal head in the future set time based on a linear regression or curve fitting prediction algorithm. According to the application, continuous, dynamic and accurate monitoring and visual display in the uterine contraction period can be realized, doctors are assisted in scientific decision making, and the delivery management level is improved.

The application belongs to the technical field of material surface structure detection, and provides a paper defect detection method and system, height information of each pixel point on the surface of the paper is acquired through an optical distance measuring sensor, and then a height distribution map is constructed, and the fluctuation mode of the surface structure contour of the paper is analyzed by combining convolution difference, local trend and change intensity, through curve fitting and trend clustering of the structure characteristics, typical surface structure types such as periodic corrugation and non-periodic wrinkle can be effectively distinguished, so that automatic identification of the microstructure of the paper surface and structure cause tracing analysis are realized, and meanwhile, accurate identification of the abnormal area is realized, and the universality and practicality of the system in the actual industrial scene are improved.

The invention discloses a point cloud-based reverse modeling curved surface generation method and system, and relates to the field of electric digital data processing, and the method comprises the steps: obtaining three-dimensional point cloud data of a to-be-processed engineering design product, and projecting the three-dimensional point cloud data to a preset cutting plane to obtain a two-dimensional projection point set; and performing dimension reduction processing on each projection point, calculating a local geometrical relationship between the projection point and a neighborhood point, and mapping the local geometrical relationship to a one-dimensional parameter space to obtain a one-dimensional parameter value. Clustering analysis is conducted on the two-dimensional projection point set based on the one-dimensional parameter values, the two-dimensional projection point set is divided into a plurality of point cloud clusters, and each point cloud cluster corresponds to one geometric feature curve. And sorting the projection points in each point cloud cluster according to a spatial position relationship to obtain an ordered point sequence. And identifying the geometric feature type of each ordered point sequence, performing curve fitting, and extracting parameter information to obtain the feature parameters of the engineering design product. And generating an accurate curved surface model according to the characteristic parameters. By implementing the method, the accuracy of point cloud clustering can be improved.

The invention discloses a power grid black-start path optimization method and device, electronic equipment and a storage medium, and the method takes the maximum total generating capacity of a system as an optimization target, and is specifically improved as follows: simply adding the actual cost and the estimated cost in a conventional A * algorithm, and improving the calculation of a cost function into the introduction of a weight coefficient. The weight coefficient is used for adjusting the ratio of the actual cost to the estimated cost, so that the efficiency and directivity of path search are enhanced, finally, curve fitting is performed on an initial black-start path output by the A * algorithm by adopting an interpolation method, and a smooth black-start path with continuous derivatives and geometric continuity is constructed, so that the problem of overlarge turning among path points is solved, and the path search efficiency and directivity are improved. And potential safety hazards such as voltage fluctuation and protection maloperation caused by path abrupt change are avoided.

The invention provides a method for performing parameter fitting on spectroscopic measurement data, and a wafer data processing method and equipment. The method comprises the following steps: acquiring target spectroscopic measurement data; performing curve fitting based on the target spectroscopic measurement data to obtain a first fitting equation; calculating a residual error between the first fitting equation and the target spectroscopic measurement data; judging whether a trough exists in the residual error or not; if a trough exists in the residual error, adding a first preset peak to the first fitting equation to obtain a new first fitting equation, and returning to carry out iteration to obtain a candidate fitting equation; calculating a first root mean square error of the candidate fitting equation and the target spectroscopic measurement data; judging whether the first root mean square error meets a preset convergence condition or not; and if the first root mean square error does not meet the preset convergence condition, adding a second preset peak to the candidate fitting equation to obtain a new candidate fitting equation, returning to carry out iteration, and taking the corresponding candidate fitting equation as a target fitting equation of the target spectroscopic measurement data based on the preset convergence condition.

The invention discloses a low-temperature environment LNG (Liquefied Natural Gas) metering sensor drift compensation method and system, and particularly relates to the technical field of sensor drift compensation. Output data of a sensor are collected and processed, and a temperature compensation model of the sensor is established; in the working temperature range, digital filtering processing is carried out on data of displacement change of each measuring range point, and then analysis and mining are carried out; a data longitudinal interval in a temperature range is refined by using a table look-up method, so that the data interval is reduced. And calculating the G value according to the residual error obtained by the table look-up method and the numerical value change condition in each temperature section. If the G value is greater than a preset threshold value of the system, selecting a proper fitting method, such as a least square method or Legendre orthogonal polynomial fitting, according to the distribution of the G value and the data interval when curve fitting is used, so as to further optimize the temperature compensation model; and if the G value is smaller than the system preset threshold value, no subsequent operation is carried out.

A manipulator execution path planning method for a complex space guided measurement task, includes: determining a measurement space of a manipulator and a starting point and a target point of a path; determining an obstructed space and a free space of the measurement space and a proportion of the obstructed space in the measurement space using a collision algorithm, and classifying and constructing a space compression model according to the proportion of the obstructed space in the measurement space to compress the measurement space and remove an invalid space; finding candidate paths for connecting the starting point and the target point by constructing an adjacent node tree connected topology network; finding a required path in the candidate paths, and performing high-order curve fitting to obtain a final path; discretizing the final path, and solving joint angles of the manipulator to obtain a joint track.

The invention relates to a road boundary line detection method and device, a vehicle and electronic equipment, and relates to the technical field of vehicles, in particular to the technical field of intelligent driving. The reference boundary line comprises a radar detection boundary line or a foresight detection boundary line; screening candidate fitting points from the radar detection points based on the reference boundary line; the transverse distance between the candidate fitting point and the reference boundary line does not exceed a distance threshold value; performing curve fitting based on the candidate fitting points to obtain a boundary fitting line; and fusing the reference boundary line and the boundary fitting line to obtain a road boundary line. Through optimization of a multi-sensor fusion technology and an environment sensing algorithm, the robustness and reliability of the intelligent driving system in a curve environment are remarkably improved, and safe and stable operation of a vehicle is ensured.

The invention discloses a method and device for compensating the dispensing height through curved surface fitting. The method comprises the steps that three-dimensional coordinate point information of a plurality of discrete sampling points on the surface of a to-be-dispensed target is collected; building a polynomial of a curved surface model for compensating dispensing according to the number of the sampling points; constructing a matrix, and constructing a linear equation set based on the primary function; and the weight coefficient of the linear equation set is solved through a Gaussian elimination method, a fitted quadric surface equation after interpolation calculation is obtained, and the dispensing height is compensated based on the quadric surface equation. According to the method, the polynomial is fitted through the three-dimensional coordinate point information, the linear equation set is constructed, the predicted compensation value is obtained after interpolation calculation is carried out according to the fitting result so as to compensate the dispensing height, the high-precision continuous curved surface model can be constructed by collecting discrete sampling points, real-time interpolation calculation of any coordinate is supported, and the precision of the dispensing height is improved. The method can solve the problem of insufficient fitting precision of dispensing height compensation based on a single plane.

A method for automatically estimating a leaf inclination angle distribution of a wheat canopy based on a voxel segmentation normal vector algorithm, including the following steps: step 1: obtaining point cloud data of a wheat canopy; step 2: splicing and denoising point clouds; step 3: calculating normal vectors of the point clouds; step 4: performing voxelization on the point clouds; step 5: segmenting the normal vectors by using a voxel; step 6: calculating angles of the voxels; step 7: collecting statistics on the angles of the voxels and performing curve fitting calculation to obtain a leaf inclination angle distribution and an average leaf inclination angle. The average leaf inclination angle estimated by the method is compared with field measured data, and the feasibility of the algorithm is verified by using a three-dimensional radiative transfer model.

A navigation path can be determined for an object using one or more neural networks. In various embodiments, image data is obtained that is representative of an environment in which the object is to be navigated. Relevant features are identified from the image, and a curve fit to those features. Loss values for the potential paths are scaled based at least in part upon the distance of those features in the real world. This can include, in at least some embodiments, performing the scaling as a function of the curvature of the curve fit to the features. Temporal smoothing can be performed with respect to prior path predictions in order to prevent sudden changes in the predicted path. The paths are analyzed to select a path with a highest confidence value that also at least satisfies a minimum confidence criterion. The path can be converted into three-dimensional navigation information.

The invention belongs to the technical field of data processing, and particularly relates to a wind turbine generator grid-connected test data processing method, device, equipment and medium under extreme weather, and the method comprises the steps: collecting a grid-connected test data set; removing abnormal data in the grid-connected test data set based on an adaptive multi-scale density clustering algorithm to obtain a first data set; and removing abnormal data in the first data set based on a Pauta criterion to obtain a processed data set. According to the method, the adaptive multi-scale density clustering algorithm is combined with the Pauta criterion, so that abnormal test data can be identified and eliminated; a cubic spline function-elastic network regularization dynamic fitting data curve is adopted, smooth curve fitting is achieved, and the method has the advantages of being high in accuracy, good in universality and the like in the aspect of wind turbine generator grid-connected test data processing under the extreme meteorological environment and the extreme meteorological environment.