16 results about "Orthogonalization" patented technology

The application belongs to optical communication system, disclose a kind of interference management method and device in optical communication system, comprising the following steps: S1, the channel state information between each transmitter and receiver in optical communication system is collected in real time, data model is constructed based on channel state information;S2, based on data model, the beam parameter of LED array is dynamically adjusted by adaptive beam forming strategy;S3, interference signal parameter in channel state information is used, and the interference signal of each transmitter is structured alignment processing by interference vector orthogonalization algorithm;S4, set interference suppression evaluation criterion, whether current interference level meets the requirement based on evaluation criterion is judged, if it does not meet the requirement, start dynamic time slot scheduling strategy;S5, the data model of channel state information is updated according to preset period, repeat steps S2-S4, beam parameter, interference orthogonalization parameter and time slot allocation parameter are continuously adjusted by iterative optimization algorithm, realize dynamic adaptive interference management.

This invention discloses a method for solving the 5-parameter coordinate transformation matrix of inertial navigation. The method includes: determining five parameters in two rows of the coordinate transformation matrix as variables for real-time updates; and using an integral formula to perform integral update calculations on the 5-parameter differential equation to obtain the five parameters at time t. k The time value is calculated; five parameters are orthogonalized and normalized; the remaining three parameters of the coordinate transformation matrix are calculated, and the updated coordinate transformation matrix is ​​obtained, supporting velocity and position updates to improve the accuracy of inertial navigation. This invention uses the angular rate output by the gyroscope orthogonally mounted on the strapdown inertial system as input information to achieve real-time updating of the inertial navigation coordinate transformation matrix, simplifying the computation while ensuring solution accuracy.

This invention relates to the field of finite element stress calculation technology, specifically to an aortic stress calculation system based on finite element analysis. In this invention, multiple sets of aortic displacement vectors and element stress vectors are spliced ​​into a response matrix according to the sequence of nodal degrees of freedom and element numbering. The Gram-Schmidt orthogonalization method is introduced to process the retained vectors, compressing the directional coincident components and linearly correlated components, resulting in stronger stability of the reduced-order basis space and smoother conversion of high-dimensional discrete equations to low-dimensional expressions. The Newton-Raphson algorithm is introduced in the path state update stage, where load step difference, nodal displacement increment, and integral point stress tensor are continuously recorded. The path turning state and load propagation form are adjusted synchronously. Facing high internal pressure, large deformation, and nonlinear response scenarios, peak tracking continuity is stronger, the frequency of solving repeated integral models is reduced, the high-risk area of ​​the aortic wall is more concentrated, and the numerical support for scheme comparison, parameter scanning, and risk pre-assessment is more sufficient.

This invention provides a simulation method and system for the processing of silicon carbide ceramic pump components based on digital twins. Specifically, it includes: generating an initial high-fidelity simulation snapshot in a preset parameter domain; constructing an event-driven weighted snapshot matrix containing a criticality function for processing events; obtaining an initial reduced-order basis vector through singular value decomposition; and establishing an initial reduced-order model through Galerkin projection. During simulation, a multi-scale thermomechanical correlation fracture sensing error index, weighted by macroscopic constitutive residuals, mesoscopic phase field gradients, and microscopic thermodynamic correlation terms, is used to judge accuracy in real time. When the error exceeds a threshold, a new high-fidelity snapshot is sampled at the current parameter point, and the basis vector is expanded after Schmitt orthogonalization and normalization with the model prediction residual vector to update the reduced-order model. The physical field response is solved in the expanded reduced-order space, and the reduced-order solution is reconstructed in high dimension through an offline-trained correction network to restore high-dimensional features, thereby realizing the simulation of the silicon carbide ceramic pump component processing.

This invention belongs to the field of hierarchical text classification and discloses a hierarchical text classification method based on label geometry construction and alignment. Based on the label set and label hierarchy structure, the embeddings of each label are obtained. Global orthogonalization is performed on the embeddings of the head labels, and recursive local orthogonalization is performed on the embeddings of parent and child labels. The orthogonalized embeddings are used as initial label embeddings. A hierarchical awareness cue fine-tuning module is trained based on the initial label embeddings and text samples. After training, the output serves as a multi-task classification model. During training, a hierarchical geometry regularization loss is calculated for hierarchical geometry alignment. The hierarchical geometry regularization loss includes a global orthogonal loss calculated based on the head label embeddings, and parent-child similarity loss, norm decay loss, and sibling residual orthogonal loss calculated based on the parent-child label embeddings. This invention can significantly improve text classification accuracy.

This invention relates to the field of intelligent control technology for the operating status of air compressor groups, and discloses a method and system for intelligent control of the operating status of air compressor groups. The method collects real-time operating data of the air compressor group, extracts energy efficiency, load, and pipeline pressure fluctuation characteristics, and generates a sequence of state feature vectors through orthogonalization processing; it uses a temporal attention network to fuse time decay and operating condition correlation, encoding and outputting hidden state vectors; it generates optimal control actions through a policy network based on deep reinforcement learning; and it combines feedforward and feedback control to generate composite control signals to drive actuators, achieving intelligent control of the air compressor group. This method improves energy efficiency, stabilizes pipeline pressure, reduces equipment start-up and shutdown frequency, and extends equipment lifespan.

The application discloses an open vocabulary segmentation method based on parallel cost aggregation and expert perception, and relates to the technical field of open vocabulary segmentation. The method performs spatial aggregation and category aggregation on a cost body in parallel, forms spatial context flow and category semantic flow, and efficiently fuses the two knowledge flows through an expert-driven perception learning module; a multi-expert parser in the module extracts complementary features from multiple perspectives, a coefficient mapper adaptively learns pixel-level weights for different features, realizes weighted integration of complementary knowledge, and obtains unified feature embedding. In addition, the application proposes a feature orthogonal decoupling strategy, which constrains the redundancy of the two features to alleviate the correlation between the semantic flow and the context flow, so that the expert-driven perception learning module can learn more diverse knowledge representation from the orthogonalized features. The application aims to alleviate the knowledge interference problem between the category-level semantics and the spatial context in the existing method, and promote more robust graph-text alignment learning.

The application relates to the technical field of image extraction, in particular to a video key frame self-adaptive extraction method based on a large model, which comprises the following steps: receiving a to-be-processed video stream and decoding the to-be-processed video stream into continuous sequence frames, performing global semantic analysis on the sequence frames to generate a global semantic text vector, and simultaneously performing pixel feature extraction on each frame image in the sequence frames. The application quantifies the expression contribution of local pictures to a global theme, suppresses the interference of noise frames which are visually clear but irrelevant in semantics, performs a greedy iterative search on this basis, calculates an orthogonal projection component in the remaining candidate set and selects a frame which can maximize the amount of new information as a semantic orthogonal key frame, utilizes an orthogonalization strategy to ensure the independence of the selected frame in a feature space and eliminate the semantic overlap between the key frames, and finally generates a self-adaptive video summary which has high information coverage and low redundancy, thereby improving the accuracy and browsing efficiency of long video content understanding.

This invention belongs to the field of educational data analysis and machine learning technology, specifically disclosing a method for analyzing performance influencing factors and causal effects that integrates Direct Machine Learning (DML) and Conformal Prediction (CP). This method introduces the orthogonalized causal estimation mechanism of dual machine learning into the analysis of performance influencing factors, and combines it with the interval construction idea of ​​conformal prediction to achieve the identification of performance influencing factors and the evaluation of causal effect intervals. By constructing outcome variable models and treatment variable models, the influence of high-dimensional covariates is eliminated through residualization, and orthogonal scores are used to achieve robust estimation of the causal effects of candidate factors on performance variables. Furthermore, the orthogonal scores are used as a measure of inconsistency in conformal prediction to determine the acceptance region of candidate causal effect parameters, thereby generating causal effect intervals. This method has good interpretability and practicality in identifying performance influencing factors, estimating causal effects, and quantifying intervals.

The application discloses a bus duct joint degradation early warning method based on machine learning and causal inference, in order to solve the problem of indirect estimation under the condition of no direct contact resistance measurement, the application realizes the robust estimation and interval early warning of the contact resistance and the degradation rate under the complex environment and load disturbance, improves the early warning accuracy and reliability by means of the event-driven neural tool variable modeling, the orthogonalization constraint, the multi-channel thermoelectric physical information neural network coded by the energy conservation and the environmental boundary, the contact resistance coupled to the current square heating term, the counterfactual gate generation baseline temperature rise and residual causal decomposition sharing the weight, and the risk score constructed by fusing the model parameters and the extrapolation uncertainty.

This invention discloses a method for obtaining beamforming weight vectors in an integrated radar-communication system, within the field of array signal processing technology. The method comprises the following steps: S1: Establishing a receiver signal model for a uniform linear array antenna in the integrated radar-communication system; S2: Introducing orthogonal polynomials and the Gauss-Legendal numerical integration method to perform a low-complexity, fast fitting solution for the interference subspace; S3: Introducing the Schmitt orthogonalization method to recursively solve the orthogonal complement projection operator of the interference subspace and project it to obtain the beamforming weight vector; S4: Introducing an equivalent representation of the beamforming weight vector under a dual-phase shifter architecture to obtain the updated beamforming weight vector value under the dual-phase shifter architecture. This beamforming weight vector acquisition method solves the problems of high computational complexity, poor performance in low-quantization-bit hardware scenarios, and desired signal suppression under high signal-to-interference-plus-noise ratios in existing methods, achieving low-complexity and highly robust beamforming.

The application relates to the fields of machine vision and intelligent welding, and discloses a fillet weld point cloud feature extraction method based on bidirectional Euclidean constraint and quartile distance statistical truncation. In non-standard flexible sheet processing, the theoretical features are suspended due to thermal deformation, the side wall topological features are misdirected, and the end points of the teaching-free system are deviated due to point solid welding splashing, pose singularity and interference collision and other composite pain points; the application is characterized in that the space affine dimension reduction is realized through the Rodriguez transformation, the side wall interference is accurately removed through the rotating minimum circumscribed rectangle feature, the entity point array deeply embedded in the gap is extracted through the bidirectional Euclidean projection joint evaluation model, the random assembly gap is resisted through the multiplicative feedback state machine, the non-parametric statistical truncation mechanism of the interquartile range (IQR) is introduced after the fitting feature, the heavy-tailed noise is adaptively stripped to accurately position the physical end point, the TCP pose is reconstructed through the Schmidt orthogonalization, the four-point anti-collision servo track is generated, and high-fidelity dynamic tracking and safe anti-collision guidance under extreme working conditions are realized.