Patents
Literature
11results about How to "Numerically stable" patented technology
Efficacy Topic
Property
Owner
Technical Advancement
Application Domain
Technology Topic
Technology Field Word
Patent Country/Region
Patent Type
Patent Status
Application Year
Inventor
Motion quality evaluation method based on video and skeleton bimodal fusion
PendingCN121767829Aimprove understandingNumerically stableCharacter and pattern recognitionBiological modelsFeature extractionBi modal
The invention discloses an action quality evaluation method based on video and skeleton bimodal fusion, and the method comprises the steps: inputting a video frame sequence containing a complete action process into a fragmentation network, and generating continuous fragments; secondly, inputting each video clip into a video modal feature extraction network and a human body skeleton feature extraction network, respectively obtaining a video modal feature and a skeleton modal feature, fusing the video modal feature and the skeleton modal feature through a cross-modal attention mechanism, and generating a fused clip-level feature representation; and finally, inputting each segment-level feature representation into a multi-layer Transform encoder, obtaining a global time sequence feature of the action through a self-attention mechanism, inputting the global time sequence feature into a multi-layer perceptron (MLP) regression network, and generating an action quality score. According to the method, a prediction result is ensured to be more stable in numerical value and more accord with human perception in semantics, and the accuracy and consistency of action quality evaluation are remarkably improved.
Owner:HANGZHOU DIANZI UNIV
A TCR Measurement Method Based on High and Low Temperature Data Fusion Using an Improved Particle Swarm Optimization Algorithm
ActiveCN121682648BSolve FragmentationSolve the problem of dimension mismatchEnsemble learningBiological modelsDigital dataDecision maker
This invention relates to the field of electrical digital data processing technology, specifically a TCR measurement method based on high and low temperature region data fusion using an improved particle swarm optimization algorithm. The method inputs the original time-series data from the high and low temperature regions, removes noise and outliers using isolated forest and Grubbs criterion, and maps the data to a unified temperature domain grid using piecewise Lagrange interpolation. An ergodic sequence is generated using Logistic chaotic mapping to initialize the particle swarm position vector, which is then transformed into node coordinate parameters in geometric space, and a cubic basis spline fitting curve is constructed. Particle fitness is evaluated using the sum of squared residuals and curvature smoothness as constraints. During iteration, the average Euclidean distance from the particle to the swarm centroid is calculated to generate a spatial dispersion factor. A multi-interval weighted decision maker based on the Mamdani framework is used to identify probe, development, convergence, and escape states, dynamically adjusting the inertia weights accordingly. This effectively solves the problems of data fragmentation and premature convergence, ultimately outputting a continuous and smooth resistance temperature coefficient distribution across the entire temperature domain.
Owner:HAINING YUNHUANG NEW MATERIALS CO LTD
ELM fault identification method for fusing feature attention and manual LU decomposition for multiple industrial devices
PendingCN121834506Aquality improvementStable and controllableBiological modelsComplex mathematical operationsNumerical stabilityIndustrial equipment
The invention discloses a multi-industrial equipment-oriented ELM fault identification method fusing feature attention and manual LU decomposition, and belongs to the technical field of industrial equipment fault diagnosis. The core of the method is that a feature attention mechanism and a manual LU decomposition module are fused: firstly, the feature contribution degree is quantified through a channel attention mechanism, key fault features are automatically focused, weak correlation feature interference is inhibited, and the quality of a hidden layer output matrix is improved; secondly, the manual LU decomposition module realizes autonomous controllability and numerical stability of matrix inverse operation through row principal component selection and regularization optimization, and is separated from the dependence of a third-party math library. According to the method, the accuracy of equipment fault identification is remarkably improved, the omission ratio and the false alarm rate are effectively reduced, memory limitation of edge equipment is overcome, and an efficient and reliable real-time identification scheme is provided for bearing abrasion and gearbox abnormity of a wind power variable pitch system and faults of other industrial equipment such as a machine tool and an industrial motor.
Owner:ANHUI UNIVERSITY OF TECHNOLOGY
Production device of glucosyl stevioside
ActiveCN224186174UImprove conversion rateResidue reductionBioreactor/fermenter combinationsCation exchanger materialsSucrose synthetaseUridine diphosphate
The utility model discloses a glucosyl stevioside production device, and relates to the technical field of glucosyl stevioside production, an inlet of a conversion tank is communicated with a raw material tank, a cyclodextrin tank, a beta cyclodextrin transferase tank and a first alkali liquor tank, and an outlet of the conversion tank is communicated with a reaction tank; an inlet of the reaction tank is respectively communicated with a sucrose tank, a sucrose synthase tank, a uridine diphosphate tank, a UGT glycosyltransferase tank and a second alkali liquor tank through pipelines, an outlet of the reaction tank is communicated with a cation exchange resin column, an inlet of the cation exchange resin column is communicated with an ethanol tank, an outlet of the cation exchange resin column is communicated with a concentration tank, and the concentration tank is communicated with a second alkali liquor tank. An outlet of the concentration tank is communicated with a dryer, and an outlet of the dryer is communicated with a glucosyl stevioside tank. The residue of unreacted glucoside in the glucosyl stevioside is reduced through a double-enzyme method, the conversion rate of a substrate is improved, the residue of the substrate is further reduced, and the taste of the product is integrally improved.
Owner:DONGTAI HAORUI BIOTECHNOLOGY CO LTD
A heuristic dynamic memory weighting and recall rearrangement method
PendingCN122087093AAchieve differential attenuation speedImprove the level of intelligenceDigital data information retrievalSemantic analysisAlgorithmTheoretical computer science
This invention discloses a heuristic dynamic memory weighting and recall reordering method, comprising six steps: initial impression value calculation, Ebbinghaus forgetting curve modeling, recall depth parameterization, multi-dimensional independent sorting, multi-dimensional result fusion, and recall-triggered impression value inversely proportional increment. This invention simulates the natural decay of memory through an 8-stage Ebbinghaus forgetting curve, achieves a recall-as-reinforcement effect through an inverse impression value increment mechanism, comprehensively considers semantic similarity, temporal freshness, and memory importance through multi-dimensional reordering, and flexibly adjusts the breadth and depth of recall through recall depth parameterization. This invention provides two forgetting curve implementation methods: discrete-stage lookup table method and continuous curve calculation method (including exponential decay type and power-law decay type), suitable for different application scenarios. This invention significantly improves the intelligence level and anthropomorphism of AI long-term memory systems, enabling AI to perform natural memory recall like humans.
Owner:LIAOCHENG DONGLI ZHIYUAN DIGITAL TECHNOLOGY CO LTD
Rotary steering dynamic tool surface measurement method and equipment based on data fusion, and storage medium
PendingCN121993146AEffectively filter out interfering noiseFilter out disturbing noiseSurveyBiological modelsQuaternionClassical mechanics
The invention discloses a data fusion-based rotation-oriented dynamic tool surface measurement method and device and a storage medium, and the method comprises the steps: obtaining an optimal gravity tool surface estimation value through an unscented Kalman filtering algorithm based on a quaternion theory of rotation coordinate conversion; based on a rotary steering tool surface calculation principle, obtaining a theoretical calculation gravity tool surface estimation value; and fusing the optimal gravity tool face estimation value and the theoretical calculation gravity tool face estimation value to obtain a final prediction value. The rotary steering dynamic tool surface is measured by combining an unscented Kalman filtering measurement result and a theoretical derivation calculation result, the numerical value is more stable, and the precision is higher.
Owner:CNPC BOHAI DRILLING ENG +1
Cylinder parallelism verticality detection device
ActiveCN224080966Uquick switchNumerically stableMeasurement devicesClassical mechanicsStructural engineering
Owner:BIKVI PRECISION IND (DONGGUAN) CO LTD
Male-female seat integrated elastic connector
ActiveCN115313079Breliable electrical connectioncontact stabilityCoupling contact membersTwo-part coupling devicesContact failureElectrical connection
The application discloses a female-male seat integrated elastic connector, which comprises a female seat assembly and a male seat assembly. The female seat assembly comprises a base and a contact terminal, and the contact terminal is installed in the base. The male seat assembly comprises a bearing plate and an equal number of elastic bodies and metal guide columns as the contact terminal. The bearing plate is fixedly installed on the upper end of the base, the metal guide columns are insertedly installed on the bearing plate, the upper end of the metal guide column is provided with a metal contact block, the lower end is elastically in contact with the contact terminal, and the elastic body is sleeved on the metal guide column and installed between the contact terminal and the bearing plate. Therefore, the metal guide column is in stable contact with the contact terminal when being inserted or vibrated, stable electrical connection can be realized, power failure or instantaneous breakage does not occur, and stable digital signals can be carried. Meanwhile, under the action of external force, the elastic body is elastically compressed and provides stable reverse elastic force to the metal contact block, the metal contact block is stably in contact with the conductive contact piece in the whole process, and therefore, poor contact or breakage of the metal contact block and the conductive contact piece is effectively avoided.
Owner:王锦炎
Modelica-based modeling method for wind wave current coupling environment of FPSO system
ActiveCN119337782BConvergence checkNumerically stableGeometric CADDesign optimisation/simulationModelicaMarine engineering
This invention discloses a modeling method for wind-wave-current coupled environment of an FPSO system based on Modelica, relating to the field of model convergence technology. By setting different time steps and running simulations, the responses of key physical quantities are recorded, and a time step sensitivity index is calculated. The model is then divided into different meshes and simulated, and a mesh independence index is calculated. The dynamic response of the FPSO platform is recorded, and an instability index of the simulation results is calculated. Combining the time step sensitivity index, mesh independence index, and simulation result instability index, a convergence anomaly coefficient is calculated and compared with a preset threshold to determine whether the model's convergence is acceptable. This process verifies the standardized convergence of the Modelica-based FPSO system wind-wave-current coupled model, ensuring timely detection of model convergence performance, avoiding error accumulation, improving the accuracy and reliability of simulation results, and thus reducing potential impacts on engineering decisions and safety assessments.
Owner:CIMC OFFSHORE ENG INST +2
Electric power equipment electromagnetic transient simulation delay compensation method and system, medium and equipment
PendingCN121995787ANumerically stableSimulator controlDesign optimisation/simulationCapacitanceTransient state
The invention provides an electromagnetic transient simulation delay compensation method and system for electric power equipment, a medium and equipment, and the method comprises the steps: obtaining a current of a controlled source equivalent circuit formed when a bridge arm in the electric power equipment introduces a controlled voltage source; wherein the controlled source equivalent circuit performs equivalent replacement on a capacitor of the capacitor equivalent circuit of the bridge arm; based on the current, calculating a circuit state characteristic root of the controlled source equivalent circuit by using a delay function corresponding to the controlled voltage source; and performing delay compensation based on the circuit state characteristic root of the controlled source equivalent circuit and the obtained circuit state characteristic root of the capacitor equivalent circuit. According to the current of the controlled source equivalent circuit and the corresponding time delay function, the circuit state characteristic root of the controlled source equivalent circuit is obtained, and then according to the circuit state characteristic root of the controlled source equivalent circuit and the circuit state characteristic root of the capacitor equivalent circuit, simulation of the controlled voltage source is controlled, and it is guaranteed that the numerical value of the power electronic equipment system is stable.
Owner:CHINA ELECTRIC POWER RESEARCH INSTITUTE CO LTD +2
Numerical method for reynolds stress turbulence model based on general time-dependent root scale
ActiveCN115438598Breduce dependencyNumerically stableSustainable transportationDesign optimisation/simulationNumerical stabilityNumerical methodology
The application relates to a numerical method, device, computer equipment and storage medium based on a general time root scale Reynolds stress turbulence model. The method comprises the following steps: obtaining a scale equation of turbulence with respect to a scale based on a relationship between a general time root scale and an omega scale according to an omega scale equation of an existing SSG / LRR-omega Reynolds stress model; coupling the scale equation and a known turbulence equation based on the SSG / LRR-omega model to obtain a model; obtaining a RANS equation set, coupling the model, obtaining a coupled equation set, constructing grid data of a turbulent flow field of an aircraft to be simulated when performing numerical simulation of the flow field of the aircraft, performing numerical solution on the coupled equation set according to the grid data, and obtaining a numerical simulation result of the turbulent flow field of the aircraft. When n is a positive integer, the wall boundary condition is strictly 0, and numerical stability can be realized when high-precision discretization or a complex structure / non-structure grid is adopted.
Owner:NAT UNIV OF DEFENSE TECH