8 results about "Gaussian radial basis function" patented technology

The invention relates to power load prediction, in particular to a dynamic hierarchical granularity model power load prediction method considering time sequence factors. According to the method, the fitting capability of periodic features, trend changes and sudden influences in load fluctuation is remarkably enhanced, and the problem of feature loss caused by neglect of a time sequence dynamic weight of an existing model can be solved. Comprising the following steps: S1, performing time sequence characteristic analysis on historical power load data, and converting an original non-stationary time sequence into a stationary time sequence through difference and logarithm transformation; s2, obtaining a cleaned power load data set; s3, based on the data cleaned in S2, selecting a Gaussian radial basis function (RBF) as a kernel function, and analyzing temperature, holiday identification and linear change trend key influence factors at the same time; s4, performing dynamic multi-level granulation on the data set, and dynamically adjusting the particle level according to the data mixing degree and the particle density; and S5, fusing the time sequence kernel function and the influence factor through a decision function, and predicting the future power load.

The invention relates to the technical field of photoelectric display, and provides an area array chromaticity data multi-point correction method based on a two-dimensional Gaussian basis RBF interpolation method, and the method comprises the steps: obtaining a first chromaticity data array of a measured surface and second chromaticity data of M correction points in the measured surface; constructing a two-dimensional Gaussian radial basis function interpolation model sharing a group of shape parameters based on the positions of the correction points; solving a weight parameter of each Gaussian basis function according to the first chrominance data array and the second chrominance data at the correction point; and performing point-by-point correction on the first chrominance data array by using the solved model. According to the method, fitting precision and smoothness of chromaticity space nonlinear change are effectively balanced through adaptive shape parameters, the chromaticity relation consistency is kept by adopting a multi-channel coupling correction mode, and correction precision and efficiency are effectively improved.

The invention relates to the field of infrared image processing, and discloses an AI automatic distance measurement method based on an infrared image model, which comprises the following steps: deploying a high-precision thermal infrared imager, and collecting multi-scene and multi-distance infrared images; preprocessing the original infrared image; establishing a data set containing a plurality of typical scenes according to the preprocessed image, and marking the image; an improved yolov5 model is adopted to train the data set, and accurate target detection frame information is obtained; sampling different environment temperature image samples and corresponding actual physical distance data to form a ranging training sample; inputting the accurate target detection frame information and the distance measurement training sample into a designed Gaussian radial basis function neural network model for training to obtain a physical distance measurement model, and compensating the error of the system by using the physical distance measurement model to obtain the distance between the target and the thermal infrared imager; the method has the beneficial effects that the ranging precision is improved, the environmental adaptability is enhanced, the real-time performance is excellent, and the deployment is simple and convenient.

The invention provides a hyperspectrum-based citrus leaf lesion diagnosis system and a hyperspectrum-based citrus leaf lesion diagnosis method, and the hyperspectrum-based citrus leaf lesion diagnosis system and the hyperspectrum-based citrus leaf lesion diagnosis method disclosed by the invention have the advantages that on the basis of analyzing hyperspectral imaging data of citrus leaves without diseases, lack of nutrients, black spots and yellow shoot; three parameters of yellow wave band reflectivity, infrared wave band slope and inflection point wavelength are used as characteristic quantities, and classification of four types of blades is realized by applying a support vector machine (RBF-SVM) classification model based on a Gaussian radial basis kernel function. The method solves the problems that when citrus tree disease information is detected through a field detection method at present, long-time observation with eyes is needed, subjective judgment of observers is depended, and misjudgment is likely to be caused; according to the present invention, the problem that the citrus tree disease information detection by using the chemical detection method needs the special person to detect by using the professional equipment so as not to accurately and rapidly detect each production stage of the citrus tree can be solved, and the citrus leaf disease can be rapidly and accurately diagnosed.

The invention discloses a single cell sequencing data analysis method based on Bayesian modeling, which is suitable for low expression gene identification and biological heterogeneity detection. The method comprises the following steps: S1, data acquisition and preprocessing; s2, establishing a Bayesian hierarchical model, and modeling technical noise and biological heterogeneity at the same time; s3, introducing a Poisson model and Poisson-negative binomial model combined strategy, and fitting and expressing a mean value and variance relationship by using a Gaussian radial basis function so as to accurately describe gene variation; s4, carrying out posterior inference by utilizing Markov chain Monte Carlo sampling to identify HVG and LVG genes; and S5, under the condition of lack of the spike-in gene, introducing a batch effect and a cross-batch sharing prior, and constructing a non-spike-in expansion model. Compared with a traditional method, the method has the advantages that the accuracy and robustness of expression variation recognition are improved, and the method is suitable for single cell data analysis of complex biological systems such as tumors and immune microenvironments.

The invention provides a wheat protein quantitative prediction method and system based on a near infrared spectrum, and relates to the technical field of wheat detection.The method comprises the steps that wheat near infrared spectrum data are obtained and preprocessed; a bidirectional gating circulation unit is used for extracting time sequence characteristics; carrying out dimension mapping through the projection layer; and inputting the projection features into a KAN model improved based on a Gaussian radial basis function to carry out nonlinear regression so as to obtain a protein content prediction value. According to the method, through a BiGRU and improved KAN mixed architecture, the method has strong sequence modeling capability and high-precision nonlinear regression capability, meanwhile, the model interpretability is enhanced through feature importance analysis, the problem that in the prior art, the generalization capability and the interpretability of the model are insufficient is effectively solved, and a reliable scheme is provided for rapid nondestructive testing of the wheat quality.

The present application relates to the field of computer vision, and more particularly to a face emotion classification system based on multi-modal image recognition, comprising a multi-modal data acquisition module, a data preprocessing module, a face detection and target detection module, a feature extraction module, a global context modeling module and an emotion classification output module, wherein the present application introduces a KAN nonlinear channel mapping mechanism after the visual Transformer feature output, utilizes the function approximation capability based on the Gaussian radial basis function to perform a channel-by-channel nonlinear transformation on the multi-scale features, realizes the mapping from the linear semantic representation to the nonlinear semantic space, and thus enhances the expression capability of the model on the micro-expression changes and the fine-grained emotion features; the present application introduces a state space modeling mechanism in the global context modeling module, and combines the KAN gating nonlinear modulation mechanism to realize the linear complexity modeling on the long-range dependency relationship, and simultaneously enhances the nonlinear expression capability in the state space.

The invention provides a blade cold and hot state deformation online measurement method based on a blade tip timing technology. The method comprises the following steps: step 1, establishing a blade tip contour deformation model; 2, based on the light spot triggering position and radius, a Gaussian radial basis function is adopted to represent a light spot; 3, setting an angle window function to restrain the Gaussian radial basis function; and 4, constructing an optimization objective function, and taking the maximum value of the blade top contour in the windowed Gaussian radial basis function. According to the invention, non-contact, high-precision and full-parameter static deformation measurement of the rotating blade in a high-temperature and high-pressure narrow space is realized.