17results about How to "Reduce computing scale" patented technology

The invention belongs to the technical field of fracture mechanics calculation, and particularly relates to a hybrid unit beam structure dynamic fracture phase field calculation method. The method comprises the following steps: determining material parameters required by the dynamic fracture phase field calculation of the hybrid unit beam structure; establishing a hybrid unit beam structure dynamic fracture phase field mathematical model according to the Lagrange action quantity principle, wherein the mathematical model comprises a displacement field control equation and a phase field evolution control equation; numerical discretization is carried out on the mathematical model, and a hybrid unit beam structure discrete model is established; and adopting a Newmark method in combination with an interlacing algorithm to carry out numerical solution on the mathematical model after numerical discretization until the displacement field and the phase field both meet convergence conditions, and obtaining a dynamic fracture phase field evolution result of the hybrid unit beam structure. The problem that an existing fracture phase field model cannot accurately simulate crack evolution of the beam structure on the thickness under the dynamic load is solved, and meanwhile the calculation cost for simulating the dynamic fracture of the beam structure is remarkably reduced.

The invention belongs to the technical field of underground water numerical simulation, and provides an underground water pollution migration simulation method and device, electronic equipment and a storage medium, and the method comprises the steps: constructing a three-dimensional octree grid, a control equation and a definite solution condition of a hydrogeological structure of a target region; carrying out discrete solution on the control equation in the three-dimensional octree grid by adopting a finite volume method taking a node as a center; constructing a post-processing approximation function and a local posterior error indicator on octree leaf subunits of the three-dimensional octree grid; determining a unit set contributed by a global error by adopting a Drfler marking strategy, then performing splitting or merging operation on leaf units, and performing grid updating by adopting a high-order interpolation or volume weighted average method; and according to a preset simulation termination condition, repeatedly executing underground water pollution migration simulation and finally performing visual display. According to the method, the simulation accuracy of water pollution migration is improved in a posterior error estimation mode.

This invention discloses a brain-inspired optimization method for sentence structure-based self-attention mechanisms, comprising the following steps: acquiring a large-scale unlabeled corpus, preprocessing the corpus text, and identifying sentence boundaries; constructing structured attention constraints based on sentence boundaries: constructing intra-sentence attention constraints for ordinary characters to allow direct attention computation between characters within a sentence, constructing intra-sentence compression constraints for sentence boundary positions to form sentence-level compressed representations, and constructing cross-sentence constraints to allow both ordinary characters and boundary positions in the current sentence to access the boundary positions of the previous sentence; applying the structured attention constraints to the self-attention modules of each layer of a pre-trained language model for training or fine-tuning based on the corpus; and performing performance testing and brain-likeness analysis on the optimized model. This invention reduces the computational scale and memory consumption of the attention matrix while retaining long-range information integration capabilities, thus lowering computational complexity.

This invention relates to the field of electronic packaging technology and discloses a multi-physics coupled packaging simulation method, including S1, macroscopic equivalent model construction, which establishes a macroscopic packaging model containing multi-physics loads and performs equivalent processing on the microscopic complex structure to efficiently obtain the overall temperature, displacement, and stress distribution; S2, key region identification, which combines macroscopic equivalent modeling with microscopic fine modeling and establishes a macro-micro bidirectional coupling mechanism to enable the macroscopic structural response and microscopic local behavior to be solved collaboratively under the same analysis framework; at the same time, combined with an adaptive iteration strategy, the microscopic model is dynamically introduced only in key high-risk areas to avoid the degree-of-freedom explosion problem caused by global fine modeling, thereby significantly reducing the overall computational scale while ensuring the accuracy of stress and strain prediction in key areas, effectively solving the problem of "difficulty in balancing accuracy and efficiency" in the prior art, and achieving a unity of high efficiency and high accuracy in multi-scale simulation.

The present application relates to a kind of ground potential discrete solving method for stray current space propagation analysis, including zoning profile according to the stratum space of study area, and in the region where current injection is strong, finer spatial division scale is used, in the part far from the region of interest, coarser division scale is used, to give consideration to calculation accuracy and solving efficiency;Stray current source is equivalently distributed to adjacent discrete unit node, and overall potential solving equation is established based on the conductive relationship between nodes;By numerically solving global discrete equation, the potential distribution and risk influence result of ground surface and the periphery of target facility are obtained.Compared with prior art, the present application can effectively reduce the calculation amount of large-scale space solving under the premise of maintaining high calculation accuracy, and is suitable for scenarios such as rail transit stray current intrusion, grounding grid potential analysis and transformer dc magnetic bias evaluation.

This disclosure provides a quantum chip wiring method, apparatus, and training method for a pairing neural network model. The wiring method includes mapping the coordinate information of components in the quantum chip from their positions to a preset reference region; grouping the starting interfaces of qubits within the reference region according to predetermined rules; grouping and sorting the terminal interfaces of the quantum chip; assigning a corresponding terminal interface group to each starting interface group based on a preset neural network model, and determining the pairing relationship between the starting interfaces and terminal interfaces within each group to output the pairing results; and performing serial wiring based on a preset algorithm to obtain the wiring path. This disclosure decouples pin pairing from path planning, automatically generates pin connection relationships using a deep learning model, simplifies path search to point-to-point wiring, significantly improves the automation and efficiency of quantum chip wiring, and can adapt to quantum chips of different sizes and layouts.

The application belongs to the technical field of nuclear reactor physics numerical calculation, and discloses a symmetric boundary processing method and system of a neutron transport characteristic line method. A geometry core model is constructed through an MOC program, boundary conditions of the geometry core model are set, the geometry core model is arranged with characteristic lines to obtain characteristic line arrangement information, symmetry of the geometry core model is set, a corresponding additional module is generated, an equivalent complete module of the additional module is generated according to the symmetry based on the additional module, boundary segment angular flux of the equivalent complete module is calculated, and the equivalent complete module is calculated according to the characteristic line arrangement information and based on the boundary segment angular flux of the equivalent complete module and a known source term distribution to obtain the equivalent complete module. The application realizes reduction of the difficulty of characteristic line arrangement of the additional module, and further improves the efficiency of neutron transport calculation.

The application discloses a scenic spot real scene three-dimensional digital twin dynamic construction method, and relates to the technical field of three-dimensional scene reconstruction and digital twin. The method comprises the following steps: constructing a global initial three-dimensional Gaussian sputtering scene representation of a scenic spot; constructing a difference perception state anchor point comprising a state coding vector, a spatial context descriptor and a semantic signature for each Gaussian cell; when a physical change occurs in a target area, collecting sparse multi-view new images and obtaining camera poses, generating an expected image using the current scene representation and calculating a state deviation score to determine a suspected change area; screening real change cells based on the difference perception state anchor point in the three-dimensional Gaussian sputtering representation space; taking original cell parameters as optimization prior values, performing local incremental optimization on a change core area and a boundary transition zone under a layered space mask, and performing boundary consistency fusion and model updating after optimization is completed. The application realizes incremental updating of a digital twin model when a local change occurs in a scenic spot.

The application discloses a kind of quick cross path scheduling methods of inter-city area customized public transport system, the shortest distance method between regions is used to build start point area end point set and end point area start point set, to generate start point area travel sequence Yo by insertion method, to generate end point area travel sequence Yd by insertion method, then the Yo and Yd are optimized by quick exchange method, finally the minimum value of start point × Yo × Yd is solved, which is the final inter-city area customized public transport system target path, with the shortest bus travel distance as the target;The method does not use probability distribution, effectively avoids the fluctuation of the final path result, according to the characteristics that the distance between nodes in different regions is much larger than the distance between nodes in the same region, the inter-regional traffic node set is established, and it is used as the end point of the start point region and the start point of the end point region, respectively, to maximize the range of optimal solution, effectively reducing the overall calculation scale.

The invention discloses a transmission and distribution cooperation-based mountainous area power grid source-grid load-storage progressive coordination planning method, system and device, and a medium, and the method comprises the following steps: generating an original operation scene through a sampling method according to historical load data and historical output data, and reducing the original operation scene into a typical operation scene; solving the source network load storage coordinated optimization model by adopting a decomposition algorithm, and obtaining an optimal investment decision of each planning stage through iteration; and performing load flow calculation on the typical operation scene based on the power grid frame parameters and the optimal investment decision, and calculating a new energy consumption rate. According to the invention, by constructing a full-voltage class coordinated optimization framework and a progressive rolling planning mechanism, the technical problems of bidirectional power flow, poor planning collaboration, insufficient uncertainty processing capability and the like caused by the access of high-proportion new energy to the mountain power grid are effectively solved.

The invention belongs to the technical field of underwater structure non-free field sound calculation, and relates to a rapid engineering calculation method for radiation noise of an underwater vehicle considering the influence of a lake bottom and a lake surface, which comprises the following steps of: intercepting a limited water area containing the underwater vehicle as a finite element calculation area, endowing impedance boundary conditions to the upper and lower surfaces of the limited water area, a perfect matching layer interface is given to the peripheral surface, a load is applied to carry out acoustic-structure interaction calculation, and the vibration velocity of the surface of the radiation surface of the underwater vehicle is obtained; step 2, establishing a boundary element shell calculation model only containing the radiation surface of the underwater vehicle, applying the vibration velocity obtained in the step 1, setting a series of mirror field points based on a mirror source method, and calculating sound pressure of all the mirror field points under a free field condition; based on the positions of the mirror image field points and the acoustic impedance of the lake bottom and the lake surface, the reflection coefficient of each mirror image sound path is calculated, the sound pressure of all the mirror image field points under the free field condition is superposed according to the corresponding reflection coefficients, and the total sound pressure considering the influence of the lake bottom and the lake surface is obtained through synthesis.

The invention discloses a vector information sensing system and method, and the system comprises at least one group of sensing units, the output end of each group of sensing units is sequentially connected with a group of storage units and a group of calculation units, and each sensing unit comprises a scalar signal distribution input end, a horizontal vector output end and a vertical vector output end. The storage unit comprises a horizontal vector information storage subunit and a vertical vector information storage subunit which are respectively used for carrying out in-situ storage on horizontal vector signals and vertical vector signals to form two storage states, and the calculation unit comprises a horizontal vector information processing subunit and a vertical vector information processing subunit which are respectively used for carrying out in-situ storage on the horizontal vector signals and the vertical vector signals to form two storage states. The storage states of the horizontal vector signal and the vertical vector signal are respectively subjected to weighted calculation and summarized at an output end to obtain a calculation result, and the calculation result is used for classification or discrimination. According to the method, the problems of low calculation efficiency, high data redundancy and the like in the processing process of distribution of scalar signals such as images and the like are solved on the hardware level.

The present application relates to the technical field of water conservancy and hydropower engineering design, and discloses a dam multi-physical field coupling design method, device, equipment and medium, the present application constructs a heterogeneous hybrid solution architecture of embedding a machine learning model and a numerical solution model which is constrained by a physical control equation, and differentiates model selection based on the difference in the rate of change of each physical field in space and time, fundamentally solving the core contradiction between the low efficiency of traditional pure numerical simulation and the lack of physical constraints in pure data-driven machine learning models which are prone to non-physical solutions; at the same time, through mutual feedback type two-way coupling iteration of the response results of each physical field, the solution efficiency and solution accuracy are simultaneously considered; on this basis, combined with the adaptive screening mechanism of the multi-working-condition differentiated response weight strategy, the design scheme can dynamically adapt to the core needs of different service working conditions, significantly improving the engineering practicability of the design scheme.

The application belongs to the technical field of computational fracture mechanics, and particularly relates to a dynamic fracture phase field calculation method for a hybrid unit beam structure. The method comprises the following steps: determining material parameters required for dynamic fracture phase field calculation of the hybrid unit beam structure; establishing a dynamic fracture phase field mathematical model for the hybrid unit beam structure according to the Lagrange action principle, wherein the mathematical model comprises a displacement field control equation and a phase field evolution control equation; performing numerical discretization on the mathematical model to establish a discrete model of the hybrid unit beam structure; performing numerical solution on the numerically-discretized mathematical model by using the Newmark method combined with the staggered algorithm until the displacement field and the phase field both satisfy a convergence condition, and obtaining dynamic fracture phase field evolution results of the hybrid unit beam structure. The application solves the problem that the existing fracture phase field model cannot accurately simulate crack evolution on the thickness of a beam structure under dynamic load, and significantly reduces the calculation cost of simulating dynamic fracture of the beam structure.

The invention discloses a simulation method and system for cross-regional diffusion of particulate matters under a dynamic boundary and a storage device, and relates to the field of environmental monitoring. Comprising the following steps: establishing a multi-region full-size three-dimensional geometric model containing a tunnel and a station; grid division is carried out on the model, and boundary changes caused by train movement are processed; the flow field is solved, and a dynamic velocity field under train movement is obtained; establishing an improved slip flux model to solve a particulate matter transport equation; setting boundary conditions of transient simulation; and according to the dynamic velocity field, solving concentration distribution and cross-regional diffusion conditions of particles with different particle sizes through a particle transport equation. According to the method, the simulation time is greatly shortened, the calculation memory is saved, the simulation accuracy of the particulate matters with different physical property characteristics is improved, and the migration behaviors of the particulate matters among different areas are accurately represented through dynamic boundary conditions.

The invention discloses a multi-sight-distance deep learning track line detection method based on global features, and the method comprises the steps: 1, installing a far-focus camera and a near-focus camera at the head or side surface of a train, obtaining track line images in different scenes in the train operation process, and constructing a data set; 2, in image preprocessing, a Sobel operator is used for calculating gradient energy to generate a focusing degree image, a decision diagram is generated through multi-scale morphological optimization operation, and smooth transition of fused image boundaries is achieved through Gaussian weighted fusion; step 3, constructing an improved ResNet18 backbone network, fusing multi-scale features, combining row selection and structure perception loss, and performing Adam optimization training; and 4, real-time track line detection is carried out. According to the invention, through multi-sight-distance image fusion and global feature extraction, technical support is provided for realizing high-precision and real-time detection of a train track line, automatic train driving, track state monitoring and other key tasks.

The invention discloses a simulation simplification method and device for an energy storage spring sealing ring and a medium, and relates to the technical field of sealing structure simulation analysis, and the method comprises the steps: obtaining geometric parameters of an energy storage spring, carrying out the radial compression simulation of the energy storage spring, and obtaining an original force-displacement curve of the energy storage spring; establishing an annular structure model based on the geometric parameters, and performing radial compression simulation on the annular structure model to obtain an equivalent force-displacement curve of the annular structure model; based on the difference degree between the original force-displacement curve and the equivalent force-displacement curve, carrying out iterative adjustment on geometric parameters of the annular structure model; when the difference degree between the original force-displacement curve and the equivalent force-displacement curve is smaller than a preset error allowable value, the annular structure model is used for replacing the energy storage spring to conduct finite element simulation analysis on the sealing performance. By adopting the method, the complexity of the finite element simulation model can be reduced and the simulation analysis efficiency can be improved on the premise of keeping the mechanical response characteristics of the energy storage spring.