12results about How to "Fast input" patented technology

The invention provides a context compression method based on query guidance. The context compression method comprises the following steps: filling query information and a context to be compressed into a dialogue template; respectively calculating first attention numerical values of a plurality of lexical elements in the context of the dialogue template; performing attention reconstruction on the context according to the first attention numerical values corresponding to the plurality of lexical elements; and compressing the context after the attention reconstruction to be within a preset target compression ratio. The invention further provides a context compression system based on query guidance, a storage medium and electronic equipment. Therefore, key information can be more efficiently reserved, a shorter context is input to the generative model, the reasoning efficiency of the generative model in a long context reasoning scene is improved, and the reasoning cost is reduced.

The invention provides an explosion discharge disaster intelligent evaluation method based on a convolutional neural network, and relates to the technical field of combustible gas explosion protection. The method comprises the following steps: acquiring an explosion discharge disaster assessment multi-dimensional data set; according to the obtained multi-dimensional data set, determining an explosion discharge disaster evaluation model based on a convolutional neural network model; wherein the convolutional neural network model is used for learning and internalizing a high-dimensional mapping relationship between a multi-source working condition parameter and a disaster intensity field in an explosion discharge physical process; acquiring multi-source working condition parameters of a to-be-evaluated environment; and inputting the multi-source working condition parameters of the environment to be evaluated into the explosion discharge disaster evaluation model to obtain the on-way distribution of the explosion discharge peak overpressure, the peak temperature and the peak wind speed, thereby realizing the efficient and accurate prediction of the disaster peak at any position on the whole propagation path after the explosion discharge.

The application relates to the technical field of intelligent generation, in particular to a unified autoregressive three-dimensional model automatic skeleton binding method and device, wherein the method comprises the following steps: converting an input three-dimensional model into point cloud data; performing sampling and normalization processing on the point cloud data to extract local and global geometric features of the three-dimensional model, and determining feature embedding of the point cloud data according to the local and global geometric features; based on the feature embedding and a preset autoregressive model, generating a skeleton tree token sequence and a discrete skin token sequence, to generate a skeleton tree and skin weights, and fusing the skeleton tree and the skin weights to obtain an automatic three-dimensional grid binding result suitable for an animation driving condition. Therefore, the problem that animation effect presentation is affected due to the fact that related technologies cannot uniformly model the skeleton and the skin weights and cannot process sparse skin weights in a more stable manner is solved.

The embodiment of the invention provides a vehicle interior material identification method and device and a robot, and relates to the technical field of vehicles. The method comprises the following steps: acquiring first spectral data of a to-be-identified material in the vehicle interior; extracting spectral features at a plurality of characteristic wavelengths and spectral texture features of a local region from the first spectral data; performing feature fusion processing on the spectral features and the local region spectral texture features to obtain fusion features; and inputting the fusion feature into the first model to obtain a material category identification result of the to-be-identified material. According to the method, the vehicle interior material identification accuracy and efficiency can be improved.

This application provides a method and related equipment for constructing a groundwater flow field based on irrotational conditions. The method includes constructing a groundwater flow model, dividing the internal space of the groundwater flow model into a computational grid comprising multiple triangular elements and multiple nodes; determining the head data of each node other than the observed nodes using a linear interpolation algorithm based on the head data of known observation nodes in the groundwater flow model; and determining the target head data of each node and the target flow velocity corresponding to each triangular element through multiple iterations based on the head data of each node. Based on the water level at observation well points and driven by the physical laws of groundwater movement, this application can quickly and automatically generate a physically reasonable flow field through efficient iterative calculations, providing a powerful new tool for groundwater simulation and analysis.

The application discloses a quantitative prediction method for atmospheric pollutant corrosion rate of metal under multi-gas coupling effect, relates to the technical field of atmospheric corrosion science and industrial protection, and comprises the following steps: step one, constructing a multi-gas corrosion rate calculation basic framework; taking the rate threshold value data of metal in a graded corrosion environment, the framework provides a unified operation platform for subsequent integration of other gases and adaptation of different metals. The quantitative prediction method for atmospheric pollutant corrosion rate of metal under multi-gas coupling effect realizes full-coverage quantitative analysis of the corrosion effects of three core atmospheric pollutants, namely Cl ‑ , H2S and NH3, effectively solves the technical difficulty that Cl ‑ cannot be independently calculated in the prior art, and realizes precise coupling and linkage calculation of Cl ‑ , H2S and NH3 by converting Cl into a dedicated corrosion contribution item.

The invention provides a three-dimensional component test method, system and device and a storage medium, a three-dimensional component model is obtained by obtaining a three-dimensional component file and rendering based on a webpage graphics library tool, and a test instruction sequence for the three-dimensional component model is generated for test operation in response to test information input of a user interface. Obtaining node test rendering data of the three-dimensional component model based on a preset test node, determining a node test result based on a comparison result of the node test rendering data and preset test boundary information, and when the test operation is executed, generating a three-dimensional component test report based on all the node test results and displaying the three-dimensional component test report; according to the method and the device, the webpage of the test process is realized through the webpage graphics library tool, the test task can be executed more conveniently, the structured test instruction sequence is generated in response to the input of the user interface, the creation and application efficiency of the test case is improved, and finally, the test result analysis accuracy is improved through automatic test result analysis based on the preset test node.

The invention discloses a collision test parameter recommendation method and device and a storage medium, and relates to the technical field of automobile collision test. The scheme comprises the following steps: acquiring a target collision motion parameter expected to be realized by a collision test trolley in a deceleration sliding table test; inputting the target collision motion parameters into a trained target prediction model, and obtaining recommended test parameters including damper target parameters; wherein the target prediction model constructs a training sample based on a simulation result of a deceleration sliding table test, and establishes a mapping relation from a motion parameter to a test parameter through a machine learning method. According to the scheme, the recommendation parameters highly matched with the target can be quickly output, the accuracy and efficiency of parameter configuration are remarkably improved, and the debugging process originally needing multiple iterations is greatly simplified.

The invention discloses an intelligent switching control system for a transformer in a power distribution area, and relates to the technical field of transformer switching control. A transformer detection terminal is arranged to periodically collect load data of the transformer in the power distribution area in a target area in a target power grid; a user side processing unit periodically collects user side load data in a target area, the user side processing unit processes the user side load data, a basic sequence and a recursive sequence are determined based on frequencies of different arrays to be verified in the user side load data in the processing process, a first compression mapping chain and a second compression mapping chain are constructed, and the first compression mapping chain and the second compression mapping chain are compressed. A mapping relation is established by using a shorter character sequence based on the basic sequence and the first and second compression mapping chains thereof, so that the capacity of user side load data is greatly compressed, the transmission rate of the user side load data in a network is further shortened, a switching strategy terminal can receive the data more quickly and input the data into a model, and the user side load data transmission efficiency is improved. And the generation rate of the commissioning and quitting strategy is accelerated.

The invention belongs to the technical field of garment reverse design, and provides an automatic flattening method and system for a three-dimensional long-sleeve mesh model.The automatic flattening method for the three-dimensional long-sleeve mesh model comprises the following steps that the three-dimensional long-sleeve mesh model is preprocessed; detecting boundaries of the three-dimensional long-sleeve mesh model, and determining double boundaries; performing double-layer traversal on the double boundaries, and calculating geodesic distance between every two vertexes in the double boundaries; sorting the geodesic distances from small to large, and screening n groups of vertex combinations with the minimum geodesic distances; converting the geodesic lines corresponding to the n groups of vertex combinations into automatic cutting lines, and cutting the three-dimensional long-sleeve mesh model; measuring the area variation of the triangles in the flattened model after cutting, determining an optimal cutting scheme and completing the flattening of the three-dimensional long-sleeve mesh model; according to the invention, full-automatic flattening is realized, the device is suitable for long sleeve models with different buckling states and plate types, and the cutting line meets the clothing process requirements.

This invention discloses a synergistic method for cold-season grazing and warm-season fertilization in alpine meadows, belonging to the field of alpine grassland ecological restoration and sustainable agricultural management technology. The method includes the following steps: using UAV remote sensing to acquire spatial distribution information of litter in the cold-season grassland, and dividing the grassland into patches with different litter loads based on this information, implementing differentiated grazing by region; using a soil sensor network buried in the grassland to monitor soil temperature and moisture content, dynamically defining the start and end times of cold-season grazing and the operational window for warm-season fertilization; during the warm season, spraying the grassland with liquid nitrogen and phosphorus fertilizer containing synergists according to the warm-season fertilization operational window. This invention deeply couples the physical disturbance of cold-season grazing with the chemical regulation of liquid nutrients in the warm season, solving problems such as slow litter decomposition, low fertilizer utilization, delayed greening, and decreased soil fertility in alpine meadows, achieving a synergistic improvement in grassland production and ecological functions.

The utility model provides an integrated water-cooling heat exchanger which comprises a fixed frame, a plurality of coolers fixedly arranged on the fixed frame, a fan assembly fixedly arranged on at least one side based on the coolers, and a wind scooper fixedly arranged between the fan assembly and the coolers. The fan assembly is combined to the cooler, compared with a single cooler, the cooling efficiency is higher, and the air cooling effect of the fan can be more concentrated due to the arrangement of the wind scooper; and on the other hand, the valve bodies used for control are arranged at the input end and the output end of the cooler, so that the input speed and the output speed of the heat exchange medium in the cooler are higher, and the use efficiency is higher.