411 results about "Fitting algorithm" patented technology

The invention discloses a BIM model automatic generation method and system based on point cloud data, and belongs to the field of building information modelling, and the transmission method comprises the steps: obtaining original point cloud data, and employing a filtering algorithm based on point cloud density adaptive adjustment to carry out the preprocessing of the point cloud data; constructing a voxel octree structure for the preprocessed point cloud data, and performing semantic classification on the point cloud; geometric modeling is carried out based on the segmented point cloud subsets, and a fitting algorithm is adopted to carry out shape completion on a point cloud area; semantic annotation is carried out on the components subjected to geometric reconstruction, a corresponding relation between component types and spatial attributes is constructed, fusion features based on a point feature histogram and a local curvature are adopted, and classification is carried out; a standard BIM component family is converted, and a three-dimensional BIM model is constructed through the mapping relation. According to the method, the voxel octree data structure and the deep semantic segmentation neural network model are combined, division and semantic recognition are performed on the point cloud data, the intelligent degree of the model is improved, and manual intervention is reduced.

The invention discloses a real-time pressure compensation method for a multi-axis linkage precision press, which relates to the technical field of control of a press, and comprises the following steps of: when a coefficient is detected to exceed a warning threshold value, executing abnormal identification and generating abnormal axis information; a short-stroke reciprocating test is carried out on an abnormal shaft under a safe working condition, force-displacement and force-speed curve characteristics are recorded through high-frequency sampling, and friction model parameters are updated by utilizing a self-learning fitting algorithm. And calculating a life degradation index by comparing a historical parameter sequence with the updated friction model parameters, and formulating a pre-maintenance strategy including a replacement period, spare part requirements and priorities. If the pre-maintenance strategy is not executed in time and it is monitored that the friction increasing coefficient continues to rise or the compensation amount is close to the upper limit, real-time compensation is applied, and the speed-limiting and load-limiting or emergency shutdown process is triggered. Through a multi-level safety control strategy, sudden friction abnormity can be accurately captured, and reliability and maintainability are remarkably improved.

The invention discloses an optical frequency domain reflection strain sensing method based on the combination of phase demodulation and cross-correlation, and the method comprises the steps: collecting backward Rayleigh scattering signals before and after the strain of a to-be-measured optical fiber, and extracting the phase information of a reference signal and a measurement signal through Fourier transform; a cross-correlation algorithm is combined to match a reference signal and a measurement signal at the same position of the optical fiber, and a dislocation phenomenon on a distance domain is corrected; calculating the relative phase of the matched reference signal and measurement signal at the same position, and performing phase unwrapping; distinguishing a fiber grating area and a non-grating area, and performing continuous correction on phase data by adopting a linear regression fitting algorithm so as to eliminate jump errors; on the basis of the differential relative phase, a window where the sudden change position is located is determined according to the window width obtained in the cross-correlation algorithm, segmented assignment is conducted on the strain of the window, and finally complete strain field distribution data are output. According to the method, the stability, the anti-interference capability and the measurement reliability of the phase demodulation method are improved.

The invention belongs to the application of a computer technology in the field of shot blasting processing, and discloses a shot blasting strength calculation method and system based on control parameters of intelligent shot blasting equipment. According to the method, an intelligent calculation method system is constructed by establishing a prediction model of the control parameters and the shot blasting strength. According to the method, nozzle pressure, moving speed, spraying angle, shot flow and target distance parameters of shot blasting equipment are collected in real time, workpiece material attribute data are combined, a multi-dimensional data fitting algorithm is adopted for processing, and an accurate shot blasting strength value is output. Meanwhile, a self-adaptive parameter adjusting module is developed, and dynamic calculation optimization can be carried out according to residual stress field distribution and coverage rate requirements. Compared with a traditional Almen test piece method, the method achieves real-time calculation of shot blasting strength and closed-loop control of technological parameters, the calculation error rate is reduced to be within 5%, test piece consumption is reduced by 80% or above, machining efficiency and quality stability are remarkably improved, and an innovative solution is provided for intelligent manufacturing in the field of surface strengthening.

The invention discloses a tower detection data optimization analysis method and system based on a point cloud model, and belongs to the technical field of electric power facility detection. Clustering analysis is carried out on original point cloud data of a tower by adopting a clustering algorithm, and different point cloud clusters are generated for different parts of the tower; finding out the central axis of the pole tower through a central axis fitting algorithm, carrying out the positioning and structural analysis of the pole tower according to the central axis, carrying out the sorting and splicing of the point cloud clusters of all parts of the pole tower according to the central axis, and comparing the point cloud data with ideal pole tower model data, and obtaining a target pole tower three-dimensional point cloud model; the health condition of the pole tower is evaluated through the model, integration of point cloud data of different parts of the pole tower is achieved, the construction precision of the pole tower three-dimensional point cloud model is improved, and therefore the pole tower condition detection accuracy is improved.

The invention discloses an automobile seat framework machining defect detection method based on machine vision, and particularly relates to the technical field of defect detection. The method comprises the following steps: constructing a multi-angle image acquisition and edge reflection modeling module aiming at complex defects such as weld joint pseudo soldering, microcracks, hole site deviation and collapse deformation, extracting weld joint continuity, edge integrity and hole site geometric consistency characteristics by using a deep neural network, and generating a structural feature vector; defect type recognition and credibility scoring are completed through small sample anomaly modeling and Gaussian mixture model classification, sub-pixel-level coordinate labeling of defect positions is achieved in combination with a Gaussian fitting algorithm, a defect positioning map is output, traceability analysis and severity grading are conducted based on historical data comparison, and the defect positioning accuracy is improved. The method is suitable for industrial online detection and quality closed-loop control.

The invention relates to the technical field of settlement observation, in particular to a non-contact monocular vision high-precision automatic settlement monitoring method, which comprises the following steps: firstly, performing target monitoring by using a YOLOv10s model, improving the real-time reasoning speed and recognition precision of target recognition, and solving the target recognition problem under a complex background; secondly, a digital image processing technology and an edge-based least square ellipse fitting algorithm are combined to realize accurate positioning of the target; thirdly, deriving based on a camera imaging principle, and introducing an improved world coordinate calculation method to solve world coordinates corresponding to the pixel coordinates so as to obtain three-dimensional world coordinates; and finally, calculating a settlement value of the monitoring area by taking the first frame image as a reference, predicting a future settlement value by using a deep learning model, and making an early warning for structure safety in advance. According to the invention, the installation demand and the calculation overhead are reduced, and the settlement monitoring precision is improved.

The invention relates to the technical field of concentration calculation, in particular to a water suspended matter concentration calculation method based on multi-wavelength detection, which comprises the following steps of: adopting a multi-spectral measurement method, utilizing light beams with different wavelengths to penetrate through a water body, and identifying particle size distribution and morphological characteristics of suspended matter particles; separating and correcting the turbidity contribution of the particles with different particle sizes through a least square fitting algorithm; decomposing the measured optical signal into a basic scattering component, a particle size dependent component and a background signal component by adopting a multi-level signal decoupling algorithm; and in combination with data after particle characteristic correction and optical signal decoupling, constructing a suspended matter concentration calculation model based on particle dynamics, calculating the water suspended matter concentration by using the dynamic model, and outputting suspended matter concentration data. According to the method, dynamic modeling is carried out on the concentration of the suspended matter, a dynamic equilibrium equation is solved through numerical values, and the concentration change of the suspended matter in time and depth is accurately described.

The invention discloses a current transformer distortion current reconstruction method based on a U-I trajectory curve, and the method comprises the following steps: carrying out the synchronous sampling of a voltage value and a current value at a measurement point, and carrying out the per-unit processing; constructing a coordinate system taking the current value after per unit processing as a horizontal axis and the voltage value as a longitudinal axis in the two-dimensional coordinate, and drawing a U-I track curve; analyzing U-I track curve characteristics of the current transformer in a normal operation state, a transient state and a steady state; linear transmission sections of transient saturation and steady saturation are identified; and taking sampling points in the linear transmission section as input values, restoring a U-I trajectory curve by using an ellipse fitting algorithm, extracting current values corresponding to abscissas on the restored U-I trajectory curve, and arranging the current values according to a time sequence to realize current reconstruction. According to the method, the reconstruction of the CT secondary distortion current is realized through a least square ellipse fitting method by utilizing the similarity between a U-I trajectory curve and an elliptic curve during normal linear transmission. The problem that the differential protection adaptability is reduced or even fails due to distortion current is effectively prevented, and the protection stability is effectively improved.

The invention provides a non-training burr detection method and device for a cover plate inner hole, relates to the field of burr detection, and solves the technical problem of non-target area interference caused by incapability of effectively focusing on an inner hole area. The method comprises the following steps: acquiring a current frame image of a target cover plate, and storing the current frame image as a grey-scale map; performing threshold segmentation on the grey-scale map, and extracting a maximum outer contour region in the grey-scale map; and carrying out convex hull calculation, and removing a concave part in the maximum outer contour region to obtain a regular convex hull contour. And generating an optimal ellipse boundary through an ellipse fitting algorithm. And carrying out segmentation fitting to remove outliers to obtain a final ellipse boundary. And generating a binary mask based on the ellipse fitting region, extracting a pixel set of which the gray value is zero in the mask, and outputting a burr region. And calculating the total defect area, comparing the total defect area with a dynamically adjusted area threshold value, judging whether the cover plate is qualified or not, and feeding back a judgment result to the detection control system. The device is used in the burr detection process of the inner hole of the cover plate.

The invention provides a data processing method and system for an aerosol analyzer, and belongs to the technical field of data processing of aerosol analyzers, and the method comprises the following steps: collecting light absorption signals of aerosol particles according to a light source system and a detector, and generating an original electric signal; preprocessing the original electric signal to obtain preprocessed data; extracting an absorption coefficient, a multiband light attenuation characteristic and a mass concentration parameter of the black carbon aerosol from the preprocessed data; analyzing the features based on a machine learning model, outputting a real-time prediction result of the concentration of the black carbon aerosol, and preprocessing the original electric signal to obtain preprocessed data, including the step of noise removal. According to the method, baseline drift is dynamically corrected through a polynomial fitting algorithm, wavelet transform denoising and a machine learning model are combined, the accuracy and real-time performance of black carbon aerosol concentration detection are remarkably improved, and the problems that in a traditional method, manual intervention is much, and the model generalization ability is poor are solved.

The invention provides an urban special-shaped building monomer reconstruction method based on inclination model data, and the method comprises the steps: carrying out the segmentation and recognition of a monomer building based on deep learning, European clustering and a graph cutting method, fitting a feature plane of the monomer building based on a point cloud plane fitting algorithm, recognizing a complex special-shaped building according to the normal vector gradient distribution of the feature plane, and carrying out the reconstruction of the complex special-shaped building. According to the method, the three-dimensional Mesh model of the special-shaped building is simplified, textures are filled and the like, reconstruction of a complex special-shaped building model is achieved, and the problems that an existing mainstream urban building monomer reconstruction algorithm is only suitable for common regular buildings, and automatic monomer segmentation and three-dimensional structured reconstruction of the special-shaped building cannot be achieved are solved. And the automation rate and the efficiency of reconstruction of the special-shaped building in the urban scene are obviously improved.

The invention provides a welding seam identification and trajectory optimization method and system based on three-dimensional vision, and the method specifically comprises the following steps: firstly, collecting the three-dimensional point cloud data of a welding seam, extracting the local geometric features of the welding seam, and carrying out the feature enhancement of the local geometric features, so as to screen out representative effective edge points; and then, collecting multi-source point cloud data of the welding seam area, constructing a multi-scale welding seam feature recognition network, inputting effective edge points obtained by screening into the network, realizing accurate recognition of the welding seam area, and further extracting a linear welding seam area. According to the recognition result, a fitting algorithm is adopted for conducting curve fitting on the weld joint area, and a preliminary welding track is generated; and for the condition that multiple sections of welding seams are intersected, an optimal fitting point searching mechanism is further introduced, the preliminary track is optimized, and the continuity and integrity of a welding path are ensured. And finally, based on the center line of the welding seam, the posture of the welding gun is precisely planned in combination with the dihedral structure model, and therefore high-precision track control and posture adjustment are achieved.

The invention discloses a steel pipe diameter and out-of-roundness measuring method based on ellipse fitting. The method comprises the following steps: acquiring section contour data of a steel pipe; utilizing an improved ellipse fitting algorithm to obtain an optimal ellipse model and a corresponding local point; reconstructing steel pipe section contour measurement data through a linear interpolation algorithm; and the maximum diameter and the minimum diameter of the steel pipe are calculated by using the reconstructed steel pipe section contour measurement data, so that the out-of-roundness of the steel pipe is determined. On the basis of geometric distribution of data points and high efficiency of an improved ellipse fitting algorithm, steel pipe section contour measurement data can be accurately reconstructed; according to the improved ellipse fitting algorithm, the confocal hyperbolic distance is used for replacing the algebraic distance, the actual geometric distance from the steel pipe section contour data points to the fitting ellipse can be accurately reflected, the fitting ellipse better conforms to the actual section contour of the steel pipe, effective data points are obtained, and the fitting precision is improved. Noise points and outliers generated by a measuring instrument and the environment are eliminated, and the measuring precision is improved.

The invention provides a method for extracting equivalent parameters of a cavity resonant wireless power transmission system. The method comprises the following steps: S1, acquiring system input impedance spectrum data; s2, taking the input impedance spectrum data of the system as input, and utilizing a trained LSTM regression model to obtain a preliminary predicted value of an equivalent parameter of the system; s3, based on the preliminary prediction value of the equivalent parameters of the system, performing regression fitting by adopting a nonlinear fitting algorithm to obtain an accurate value of the equivalent parameters of the system; the system equivalent parameters comprise cavity internal resistance and cavity equivalent inductance of the resonant cavity and a mutual inductance value between the coil and the resonant cavity. The method has the advantages that equivalent parameters of the cavity resonant wireless power transmission system in all current mainstream modes can be effectively extracted; the NLS algorithm is prevented from convergence failure or falling into local optimum due to the fact that an initial value seriously deviates from a real value, and the fitting precision and stability are extremely high.

The invention provides an automobile air conditioner self-adaptive adjusting method and system based on multi-source data fusion, and belongs to the field of automobile heat management control systems. The method comprises the steps that the setting state of a user for the air conditioner, environment data and user state information are obtained, and the user state information comprises user characteristics, mental states and health indexes; on the basis of the setting state, the environment data and the user state information, personalized target air conditioner parameters are generated through statistical analysis and a fitting algorithm; according to the real-time environment change, the user state and the health data, the target air conditioner parameters are dynamically compensated, and optimized air conditioner control parameters are generated; and according to the optimized air conditioner control parameters, the air conditioner operation state is adjusted. The purpose of automatically adjusting the air conditioner to the expected state of the user under different states of different users can be achieved, and precise control over the climate in the cabin is achieved.

The invention provides a self-adaptive fitting method and system for a substrate deformation measurement system based on a fiber bragg grating (FBG), and the method comprises the steps: reading an output signal of an FBG sensor network, respectively collecting the wavelength data of a grating sensor and the stress data of a stress sensor at the same position as the grating sensor, and constructing a stress-wavelength data set; fBG data points with most sensitive deformation response are automatically identified, and a dynamic sensitive point set is formed and then preprocessed; a set needing to be continuously monitored is analyzed and determined, a specific point is selected to execute an adaptive segmentation fitting algorithm, segmentation points and fitting polynomial coefficients of all subintervals are output, and a final fitting model is constructed; according to the sub-interval of the wavelength value measured by the FBG sensor, calling a sub-interval fitting model, calculating a corresponding stress value, and further obtaining the deformation quantity of the measured object. According to the invention, the contradiction between the monitoring efficiency and precision of the traditional uniform layout in a complex stress scene is solved, and the technical path and the application target have remarkable industry uniqueness.

The invention discloses a Beidou and UWB dual-mode cooperative rail transit positioning method, and particularly relates to the technical field of fusion positioning, and the method comprises the following steps: constructing a cross-modal distortion discrimination array based on the space coverage state of two types of signals in a train path, recognizing a track mutation region, and dynamically adjusting the signal fusion weight; abrupt change points are corrected through a constraint backtracking trajectory fitting algorithm, and pseudo trajectory points are eliminated in combination with a positioning base station space topological structure; performing time sequence analysis on the signal fusion weight change, and if a frequent signal hopping trend exists, lowering the related fusion weight to the lowest tolerance threshold value to improve the track continuity and the positioning robustness; according to the method, the cross-modal distortion discrimination array is constructed, accurate identification of a track mutation region is realized, abnormal track points are effectively corrected in combination with fusion weight dynamic adjustment and a constraint backtracking track fitting algorithm, a signal jump trend is suppressed through a time sequence analysis mechanism, and the stability and positioning precision of track fusion are improved.

The invention discloses a numerical control machine tool autonomous tool setting method based on 3D visual guidance. The method comprises the following steps: firstly, acquiring three-dimensional point cloud data of a cutter and a workpiece by using a 3D visual sensor and preprocessing; then, segmenting the point clouds through a region growing algorithm, realizing the separation of the tool point clouds and the workpiece point clouds, and extracting the point clouds of each cylindrical section of the workpiece; thirdly, based on a cylinder fitting algorithm, cylinder model fitting is conducted on the segmented tool and workpiece point cloud, and geometric parameters such as the axis direction, the radius and the center coordinate are obtained; calculating a relative position relation between the tool and the workpiece according to the geometric parameters, generating a tool setting path and converting the tool setting path into an NC code; and finally, the NC code is transmitted to a grinding machine control system through a grinding machine communication module, and tool setting operation is completed. The automatic tool setting device has the advantages that the relative position relation between the workpiece and the tool can be automatically measured in the machining process, automatic and rapid tool setting is completed, and the machining quality and efficiency are guaranteed.

The invention belongs to the field of gear machining detection, and particularly relates to a cylindrical gear ghost order detection method and detection equipment based on tooth surface waviness analysis. The method comprises the following steps: firstly, acquiring tooth surface deviation data of each tooth in a gear, carrying out resampling, inclination error correction and de-modification on the tooth surface deviation data, and then combining the tooth surface deviation data into full-tooth deviation curves in four directions; and then extracting waviness data and generating a full-tooth waviness curve. After RC is carried out on a full-tooth waviness curve, on one hand, the full-tooth waviness curve is converted into a frequency domain signal for selecting a frequency range, on the other hand, spectral analysis is carried out on a waviness signal through an iterative sine wave fitting algorithm within the specified frequency range, sine wave components are extracted, and a corresponding order spectrum is generated. And finally, when a sine wave component of which the amplitude is greater than or equal to 0.02 mu m and which is in a non-integral multiple harmonic order with the tooth number exists in the order spectrum of the target gear, detecting the gear ghost order. According to the invention, the problem of lack of a high-precision and high-efficiency waviness analysis and gear ghost order detection method in the prior art is solved.

The invention discloses a non-contact drunk driving detection method and system adopting a large-linewidth tunable laser, and relates to the technical field of laser detection, and the method comprises the steps: collecting transmission light signals and corresponding electric signals under different linewidth conditions through experiments, synchronously recording experiment parameters, obtaining an original data set, carrying out the preprocessing of the original data set, and carrying out the detection of drunk driving. Generating a training set, a verification set and a test set, and training a BP (Back Propagation) neural network to obtain a pre-trained noise suppression model; the method comprises the following steps of: generating a wavelength-tunable modulated laser beam by adopting a large-linewidth tunable laser, directionally projecting the modulated laser beam into a vehicle cab, and enabling laser to penetrate through the cab space and carry alcohol gas concentration information to obtain a transmission light signal; focusing the transmission light signal to a photoelectric detector through an optical convergence device to obtain a noisy electric signal, inputting the noisy electric signal into a pre-trained noise suppression model, and outputting a de-noised electric signal; performing synchronous demodulation and harmonic analysis on the de-noised electric signal by using a phase-locked amplification detection technology, extracting a second harmonic signal amplitude associated with the modulation frequency, and determining a second harmonic signal characteristic amplitude through peak detection; according to the calibration relation between the second harmonic signal characteristic amplitude and the alcohol gas concentration, the concentration value of the alcohol gas in the cab is calculated in an inversion mode through a least square fitting algorithm, an alcohol concentration threshold value is set, the drunk driving condition of a driver is judged, a detection result is generated, and an alarm signal is output.

The invention provides a dimple quality online detection method and system based on 3D scanning and virtual assembly, and the method comprises the steps: S1, obtaining dimple hole point cloud data, segmenting the point cloud data, and obtaining conical surface point cloud data; s2, performing conical surface fitting on the conical surface point cloud data through a fitting algorithm to obtain a local point set and a cone parameter; and S3, according to the local point set and the cone parameters, pit diameter contour plane fitting is carried out, the pit diameter contour plane and the nail head are assembled in a simulated mode, and a dimple depth error value is obtained. On the basis of 3D scanning and virtual assembly, pit hole surface point cloud analysis, nail rod interference analysis and compensation functions are achieved, pit depth measurement is achieved, and compared with a monocular or binocular vision scheme, the measurement precision is higher.

The invention provides a TDLAS (tunable diode laser absorption spectroscopy) second harmonic software demodulation method, which comprises the steps of modulation signal acquisition, signal conversion, equivalent sampling and signal reconstruction, mobile fitting harmonic demodulation, signal smoothing, parabola fitting peak extraction and the like. The peak value is extracted by combining moving average smoothing and parabola fitting, high-precision and low-cost real-time measurement of the gas concentration of the TDLAS system is realized by a full-software scheme, meanwhile, the complexity of the system is reduced, and the method is suitable for dynamic measurement and analysis of the gas concentration.

The invention belongs to the technical field of semiconductor device performance testing, and particularly relates to a VCSEL epitaxial wafer resistivity distribution measuring method and system, and the method comprises the steps: constructing a three-dimensional discretization model representing an epitaxial wafer and an unknown parameter vector containing the vertical resistivity, the top transverse resistivity and the substrate leakage junction nonlinear resistance; a physical response function is defined, the vector is iterated through a discrete point fitting algorithm, so that the total residual error between the analog voltage difference output by the function and the actually measured voltage value under the cross-order-of-magnitude current is minimum, and the real vertical resistivity is extracted from the optimal parameter vector; and obtaining the vertical resistivity of each point on the two-dimensional test point grid so as to generate a distribution diagram and carry out anomaly detection. According to the invention, the problem of measurement distortion caused by the leakage defect of the substrate is solved, and the measurement accuracy of the vertical resistivity is improved.

The invention relates to the technical field of insulator measurement, in particular to an insulator appearance parameter detection method and system based on RGB point cloud data. According to the method, denoising is performed on an original RGB point cloud image set of the side surface of a target insulator through a bilateral filtering algorithm, feature extraction is performed through an FPFH algorithm, and point cloud registration processing is performed through an ICP algorithm, so that a high-quality side surface global point cloud image is obtained; then, carrying out classification processing on the side global point cloud image by adopting a PointNet algorithm, and only retaining point clouds belonging to the insulated blade part; performing curved surface reconstruction on the point cloud of the insulating blade part through a B-spline curve curved surface fitting algorithm, and further combining into a three-dimensional model of the insulating blade part; and finally, projecting the three-dimensional model into a two-dimensional point cloud, calculating an edge point cloud connection line based on the two-dimensional point cloud, and further calculating a creepage distance and an overall structure distance. Based on various algorithms, the appearance parameters of the insulator can be automatically calculated.

The invention relates to a material loading control method, which comprises the following steps of: acquiring volume data of loading materials in a time period, calculating volume according to the acquired volume data, and calculating a mapping relation by utilizing a fitting algorithm; installing a camera device, capturing a material flow image in real time, and calculating a fluidity parameter in a flow velocity vector field by using an algorithm; substituting the calculated fluidity parameters into the mapping relation, correcting the mapping relation, and establishing a corrected volume mapping relation model; the prediction result is compared with the volume of the material monitored in real time, the control parameters are dynamically adjusted according to the comparison result, and the deviation in the loading process can be found and corrected in time by monitoring the volume and fluidity parameters of the material in real time and dynamically adjusting the control parameters according to the corrected volume mapping relation model; and the control precision of material loading is improved.

The invention discloses an automatic reading method and device for a pointer instrument, equipment and a storage medium. The method comprises the following steps: acquiring an image of a to-be-measured instrument, and preprocessing image data; inputting the preprocessed image into an improved lightweight YOLOv5 target detection network, detecting and positioning an instrument area, performing tilt correction on an instrument based on a prediction frame, and extracting a front view table image; performing graying processing on the front view chart disc image, and segmenting a pointer region by using an improved GA-Otsu algorithm; and extracting a pointer center line in the pointer area through an RANSAC straight line fitting algorithm, and calculating instrument reading by combining range information and a horizontal angle method. According to the technical scheme, through the lightweight network and the optimization algorithm, the instrument detection precision and efficiency are remarkably improved, and the calculation amount is reduced. The anti-interference capability is enhanced by self-adaptive filtering and tilt correction, the reading error of GA-Otsu combined with RANSAC is relatively low, the end-to-end design adapts to embedded equipment, and the labor dependence and cost are reduced.

The invention provides a conveyor belt damage detection method and system, and relates to the field of artificial intelligence, and the method comprises the steps: collecting conveyor belt damage images, and making a data set for training; based on the YOLO model, utilizing a C2PSAACmix module to enhance the extraction capability of global features and local features of the YOLO model, and utilizing an iRMB module to optimize an information flow transmission mechanism of the YOLO model so as to obtain a conveyor belt damage detection model; inputting a to-be-detected conveyer belt image into the trained conveyer belt damage detection model to obtain a segmentation result; based on the segmentation result, determining that the damage category is epidermis wear, hole or fracture; judging the fracture direction of the conveying belt by using an ellipse fitting algorithm; and determining a damage area and damage pixels according to the detection frame of the segmentation result so as to carry out quantitative analysis, and determining the damage degree and / or the fracture expansion trend of the conveyor belt according to the result of the quantitative analysis. According to the invention, different types of conveyor belt damages can be automatically identified and evaluated, and the possibility of manual intervention and operation errors is reduced.

The invention discloses a method and device for predicting average arterial pressure based on a full-connection neural network, and relates to the technical field of biomedicine based on artificial intelligence. The method comprises the steps that blood pressure pulse wave signals of each patient are collected, and S-G filtering processing is conducted on the blood pressure pulse wave signals; constructing a blood pressure pulse wave envelope line through a double-Gaussian fitting algorithm, and estimating the average arterial pressure of each patient according to the blood pressure pulse wave envelope line; respectively calculating relevancy between the multiple human body physiological parameters of each patient and the average arterial pressure; the human physiological parameter variables with the relevancy larger than a preset value serve as the input quantity of a full-connection neural network, the predicted value of the average arterial pressure serves as the output quantity of the full-connection neural network, a training sample set and a verification sample set are constructed, and the full-connection neural network is trained; and predicting the average arterial pressure of the patient by using the trained full-connection neural network. The accuracy of the non-invasive blood pressure measuring equipment can be improved.

The present invention relates to a robot spraying quality analysis method based on an improved double-β paint film distribution model. A double-β paint film distribution model is created based on a single-β paint film distribution model. The spraying speed is introduced into the double-β paint film distribution model to obtain a dynamic film thickness distribution model. The spraying height and ambient humidity are introduced. The dynamic film thickness distribution model under the introduction of the spraying height and ambient humidity is introduced. The spray width pressure, atomization pressure, and injection pressure are introduced to obtain a dynamic film thickness distribution model under multiple process parameters. The least squares method is used to fit the undetermined parameters in the final dynamic film thickness distribution model under multiple process parameters. The accuracy and generalization performance of the final dynamic film thickness distribution model under multiple process parameters meet the preset requirements and output a film thickness distribution equation. Otherwise, the fitting algorithm parameter values are adjusted and the corresponding parameters are re-derived. The dynamic film thickness distribution model under multiple process parameters is used to achieve quality analysis and prediction of the average film thickness and the standard deviation of the paint film thickness of the robot spraying process.