153 results about "Pattern processing" patented technology

The invention discloses a three-dimensional curved surface pattern layering processing method based on parameterized segmentation, and aims to solve the problem of high-precision pattern processing on the surface of a curved surface by using a laser scanning galvanometer. The method comprises the following steps: firstly, importing a three-dimensional model, and carrying out parameterization processing to obtain a two-dimensional parameter domain; positioning a to-be-processed plane pattern in the two-dimensional parameter domain to form a complete two-dimensional geometric grid covering a pattern area, generating a three-dimensional pattern grid attached to a curved surface, and calculating a center normal vector of the three-dimensional pattern grid; then, generating an equidistant plane group on the surface of the curved surface, and calculating an intersection line of the plane and the three-dimensional pattern grid as a laser scanning track; and the five-axis linkage platform-galvanometer system is started for machining. Through the parameterization domain, distortion of curved surface pattern machining is effectively solved, machining uniformity is improved, and the method is particularly suitable for laser machining of large-scale or high-complexity curved surface patterns.

The invention discloses a large language model enhanced table question answering method based on thinking chain reasoning, and belongs to the technical field of natural language processing. The method comprises the steps that a form question and answer CoTSQL model based on a thinking chain and a framework extraction model are constructed, an example screening module generates an SQL framework architecture of a natural language question through the framework extraction model, and an example similar to a target question is selected by using a screening algorithm according to a framework category; mode linking is carried out in the mode processing module, key table information and mode relations are extracted, and a context environment is provided; the prompt construction module integrates the screened examples and the processed mode information into a CoTSQL style prompt, and inputs a large language model to generate a preliminary SQL query; and the SQL correction module detects and repairs illusion errors generated by the large language model through SQL repairs and a consistency strategy, and executes revised SQL query in a database to finally obtain an accurate table question and answer result. According to the method, the accuracy and robustness of SQL query are effectively improved.

The invention discloses a marine frequency converter intelligent cooling system based on integrated modularization, relates to the technical field of ship power electronic equipment heat dissipation, and discloses the marine frequency converter intelligent cooling system based on integrated modularization. Comprising five core modules, wherein a data acquisition module is responsible for acquiring and preprocessing temperature, humidity and salt mist concentration data; the route analysis and model establishment module establishes and optimizes a model for predicting future environmental parameters by analyzing a route and applying a random forest algorithm; the threshold value calculation and optimization module calculates a dynamic threshold value in real time and continuously adjusts according to the ship speed and environmental parameters to continuously optimize threshold value parameters; the mode processing and conversion module identifies an extreme environment combination, triggers corresponding processing measures, and smoothly adjusts processing parameters and modes; and the route adjustment and safety protection module sets an intermediate value and an adjustment curve, continuously monitors the state, and automatically gives an alarm and takes measures when a fault or an abnormality occurs.

Various methods, apparatuses, systems, and media for implementing endpoint detection and response and data protection correlation are disclosed. A correlation engine receives a data stream from multiple sources and one or more patterns. A processor analyzes the data to relate an event to the one or more patterns and executes a policy when the event matches the one or more patterns to identify suspected encrypted data files. The processor transfers the suspected encrypted data files and an encryption key to a controlled testing environment to use the encryption key to safely decrypt the suspected encrypted data files and to test the suspected encrypted data files to determine a potential network impact of executing each of the suspected encrypted files on the network. The processor detonates, from the controlled testing environment, the suspected data files confirmed during the testing to have a negative impact on the network.

A data processing system and method provides automated extraction, translation, load (ETL) operations. A server coupled with the processor executes the instructions and extracts data from various sources, such as cloud storage, external APIs and direct uploads. The system may include a stream mode processing unit to process large data files in manageable blocks, thereby improving efficiency and reducing memory load. It performs integrity checks to ensure data accuracy and consistency. The system configures and applies predefined and custom transitions, which is facilitated by user-friendly interface and API integration. Custom conversion logic is integrated into the process, allowing adaptive data operations to be performed. The converted data is then verified and formatted to be loaded into different destination systems. The ETL process is efficient, extensible and user-friendly, and thus suitable for wide data processing applications.

A system includes a memory configured to store a software application and a plurality of microservices including a set of independent software service components configured to be independently executed. The system includes processors operably coupled to the memory and configured to receive a request to initiate an execution of interactions with the software application for satisfying the request, and to identify, based on the request, one or more microservices to be executed to initiate the execution of the interactions with the software application. The processors further execute one or more generative machine-learning models trained to identify anomalous interservice interaction patterns between the one or more microservices. In response to identifying the anomalous interservice interaction patterns, the processors dynamically update operations processes associated with the one or more microservices, and further execute the one or more microservices to initiate the execution of the one or more interactions with the software application.

Data processing systems and methods provide for automated Extract, Transform, Load (ETL) operations. A server, coupled with a processor, executes instructions to extract data from various sources such as cloud storage, external APIs, and direct uploads. The system can include a stream mode processing unit for handling large data files in manageable chunks, thereby enhancing efficiency and reducing memory load. It performs integrity checks to ensure data accuracy and consistency. The system configures and applies both predefined and custom transformations, facilitated through a user-friendly interface and API integration. Custom transformation logic is integrated into the process, allowing for adaptable data manipulation. The transformed data is then validated and formatted for loading into diverse destination systems. This ETL process is efficient, scalable, and user-friendly, making it suitable for a wide range of data processing applications.

A processor may capture network packets from a communication channel between a user and an application, the network packets comprising a data structure package. A processor may identify a schema of a payload used in the data structure package of the network packets. A processor may dynamically generate payload schema processing code using a machine learning dynamic protocol parser by applying a machine learning model to the schema of the payload, the payload schema processing code including a description of each field of the data structure package and a parser function for extracting a prompt. A processor may iterate the dynamically generating of the payload schema processing code until the payload schema processing code meets a predefined accuracy and functionality criteria that successfully extracts the prompt.

The invention belongs to the technical field of electronic manufacturing and printed circuit boards, and particularly relates to a lamination method and system of a multilayer flexible circuit board, which comprises the steps of providing a flexible substrate layer, forming a graphical photosensitive layer, arranging a binder layer, stacking a multilayer structure, performing exposure development and lamination treatment and the like. The system comprises a material supply unit, a graphical photosensitive layer forming unit, a lamination stacking unit, a graphical processing unit and a lamination unit, and full-process automation is realized through a central control unit. According to the invention, the positioning and alignment precision can be improved, the material compatibility is optimized, the process integration level is enhanced, and the lamination quality is improved, so that the production efficiency and the product yield of the multi-layer flexible circuit board are remarkably improved.

The invention relates to the technical field of biological recognition, in particular to a multi-scene-oriented dynamic voiceprint recognition updating method, which comprises the following steps: synchronously acquiring voice signals through a 16-channel MEMS microphone array and a throat three-axis acceleration sensor, and constructing a three-dimensional data set containing acoustic scene fingerprints, semantic contexts and emotional characteristics; extracting features by adopting a scene adaptive dual-mode processing architecture, and extracting semantic features and emotional state coding vectors to form a multi-dimensional feature set; based on the multi-dimensional feature set, calculating a matching degree by using a Gaussian mixture model, monitoring a scene mutation index, and triggering an online clustering reconstruction mechanism to optimize a feature weight; deploying a three-level pipeline architecture for real-time processing; and generating a scene adaptation factor according to the MFCC-PLP mixed feature matrix and the environmental parameters to perform multi-feature joint matching. According to the invention, the adaptability of voiceprint recognition in a complex scene can be improved.

Matrix multiplication operations can be implemented, at least in part, on one or more tensor cores of a parallel processing unit. An efficiency of the matrix multiplication operations can be improved in cases where one of the input operands or the output operand of the matrix multiplication operation is a square matrix having a triangular data pattern. In such cases, the number of computations performed by the tensor cores of the parallel processing unit can be reduced by dropping computations and / or masking out elements of the square matrix input operand on one side of the main diagonal of the square matrix. In other cases where the output operand exhibits the triangular data pattern, computations can be dropped or masked out for the invalid side of the main diagonal of the square matrix. In an embodiment, a library implementing the matrix multiplication operations is provided.

The utility model relates to a conveying linkage mechanism which comprises a working table, a conveying disc, a conveying driver, a rotating carrier, a transmission assembly, a driving assembly and a butt joint assembly, the conveying disc is rotatably arranged on the working table, the conveying driver is fixedly connected with the working table, the conveying disc is in transmission connection with the conveying driver, the rotating carrier is rotatably arranged on the conveying disc, and the driving assembly is fixedly connected with the rotating carrier. The transmission assembly is connected with the rotary carrier and comprises a transmission gear, the driving assembly is arranged on the workbench in a sliding mode through the butt joint assembly and comprises a driving gear, and in the CD grain machining process, the driving gear is driven by the butt joint assembly to move towards the conveying disc, and the driving gear is meshed with the transmission gear so as to drive the rotary carrier to rotate. According to the utility model, the workpiece can be fixed and conveyed, the workpiece can be driven to rotate at a high speed in the machining process, and contribution is made to the design of a full-automatic CD grain machining device.

The application discloses an exposure system and a photoetching machine. The exposure system comprises a light source, a micromirror array and a microlens array. The light source is used for emitting light. The micromirror array is used for receiving the light and converting the light into a first light spot array. The microlens array and the light source are respectively located on opposite sides of the micromirror array. The microlens array is used for receiving the first light spot array and performing focusing processing on the first light spot array to form a second light spot array. The microlens array projects the second light spot array onto a substrate. The photoetching machine comprises the exposure system. The micromirror array performs patterning processing on all the light and forms the first light spot array. The microlens array performs focusing processing on the first light spot array to form the second light spot array. Each microlens in the microlens array corresponds to each micromirror in the micromirror array. Compared with a traditional exposure system, the alignment accuracy problem between the microlens array and the DMD micromirror does not need to be considered.

The application discloses an exposure method and an exposure device, and belongs to the field of printed circuit board production. The exposure method of the application processes the sub-picture data and the alignment process in parallel, effectively shortening the exposure time. The application can cache the processed picture data in the host computer or the exposure system. Meanwhile, the suction cup of the exposure device of the application has the freedom of four directions of X-axis, Y-axis, Z-axis and rotating shaft, can correct the deviation of the PCB plate through rotation and translation, thereby replacing the rotation and translation processing of the pattern according to the alignment result of the substrate mark point in the sub-picture system. The process reduces the time consumption of the pattern processing in the sub-picture system, further improves the equipment capacity, avoids the step-shaped deviation problem caused by the small-angle inclined line of the horizontal line after the pattern processing, ensures the analytical precision of the exposure pattern, and improves the exposure quality.

The invention relates to an electric power data processing method and device, equipment, a storage medium and a program product. The method is applied to an electric power trusted data space and comprises the following steps: acquiring electric power data from a database accessed by the electric power trusted data space and generating a data information table; determining first power data based on the data information table; performing data structure sorting on the first power data to obtain second power data; determining sensitive data in the second power data; determining sensitive category level information corresponding to the sensitive data; acquiring data processing requirement information corresponding to the data main body; determining a data calculation mode corresponding to the second power data according to the sensitive category level information and the data processing requirement information; and processing the second power data based on the data calculation mode corresponding to the second power data. According to the method, safe and efficient processing of the power data can be realized.

According to embodiments of this disclosure, a method, apparatus, and medium for pattern processing are provided. In the method, parameter information related to a lithography machine and a wafer in a lithography process is acquired. The lithography machine is used to form a wafer pattern on the wafer based on a target pattern. Based on the parameter information, a target state representation corresponding to the target pattern is determined from a plurality of candidate state representations indicating the electron beam scattering state. Based on the target state representation, a correction result for the target pattern is determined. In this manner, the state representation of the electron beam scattering state corresponding to the target pattern can be determined based on the parameter information of the lithography machine used and the parameter information of the wafer used for lithography, thereby simulating the lithography process more accurately. This ensures that the state representation used for lithography process simulation can more accurately describe the scattering state of the electron beam used by the lithography machine, improving the reliability of the simulation process and thus more accurately determining the correction result for the target pattern.

The application relates to the technical field of blind guiding, and particularly discloses an intelligent blind guiding instrument, which comprises a host computer internally provided with an MCU processor and a power module; a head-mounted device composed of a ring frame and a camera, a microphone and a loudspeaker mounted on the ring frame; and a blind guiding stick in Bluetooth connection with the MCU processor, wherein the MCU processor is in communication connection with a background server or a cloud server and loads an image database; the MCU processor is used for receiving image information sent by the camera and identifying and judging the image category, so as to control the blind guiding instrument to enter a voice navigation mode or a thing recognition mode; the MCU processor controls the blind guiding instrument to enter a voice control mode when detecting that the microphone sends predetermined voice content, so as to start or stop the blind guiding instrument or play predetermined audio. The above-mentioned intelligent blind guiding instrument combines the blind guiding stick with the head-mounted device, collects the images around the blind guiding stick to position a position to be guided, improves the navigation accuracy, has multiple functions, realizes one machine with multiple functions, and reduces the blind guiding cost of users.

The utility model relates to the technical field of reel processing, and discloses a reel antistatic flow line processing device which comprises a main frame, a polishing disc is arranged on the top face of the main frame, a positioning support is arranged on one side of the top of the polishing disc, the positioning support comprises a connecting frame, and the bottom end of the connecting frame is connected with a sleeve frame. A servo motor is arranged in the center of the interior of the connecting frame, a threaded rod is connected to the center of the bottom end of the servo motor, a telescopic rod is connected to the bottom end of the threaded rod, and scale marks are arranged on one side of the outer wall of the telescopic rod. According to the utility model, manual work is replaced by the adjustable telescopic rod and the clamping frame, and the lifting process can be kept accurate and controllable through the scale marks distributed on the outer wall of the telescopic rod, so that the situation that the normal product quality of the reel is influenced due to unstable flow line treatment process caused by excessive attachment of the reel clamped by the bottom clamping frame and the static electricity eliminating layer is avoided; and stable clamping is achieved through the anti-skid pieces on the inner walls of the clamping plates, and the high-precision adjustable and stable treatment process is formed.

The application belongs to the technical field of electronic manufacturing and printed circuit board, and particularly relates to a laminating method and system for a multilayer flexible circuit board, which comprises the following steps: providing a flexible substrate layer, forming a patterned photosensitive layer, arranging an adhesive layer, stacking a multilayer structure, performing exposure and development, and laminating, etc. The system comprises a material supply unit, a patterned photosensitive layer forming unit, a laminated stacking unit, a patterned processing unit and a laminating unit, and realizes full-process automation through a central control unit. The application can improve positioning and alignment accuracy, optimize material compatibility, enhance process integration, improve laminating quality, and thus significantly improve the production efficiency and product yield of the multilayer flexible circuit board.

The invention discloses a micro-nano pattern processing method, which comprises the following steps of: 1, pretreating a pattern electrode and a base material, preparing the pattern electrode matched with the outline of a micro-nano pattern to be processed, performing surface activation treatment, and performing cleaning and modification treatment on the base material; step two, functional material selection and pretreatment: selecting at least two functional materials, and performing dispersion and filtration treatment to form a stable functional material dispersion system; the method has the beneficial effects that synchronous loading of multiple functional materials is achieved through the electrostatic adsorption principle, multi-material micro-nano patterns can be prepared through one-time machining, and compared with a traditional step-by-step machining method, the machining efficiency is improved, and the machining period is remarkably shortened; the precise transfer device and the visual positioning module are adopted, precise alignment of a preformed pattern and a base material is achieved, the pattern size precision can reach the nanoscale, and the problem that in traditional multi-material machining, the alignment precision is low is solved.

The invention discloses a five-axis galvanometer light beam five-dimensional integrated error correction system based on machine vision. The system comprises a vision measurement system, a Qt software system and a computer. The vision measurement system comprises a rotary table, an industrial camera and camera control software, and the rotary table controls the machining plane to rotate; the industrial camera communicates with the computer through the camera control software to shoot and store patterns; the Qt software system comprises a pattern processing module, an error measurement module and an error correction module; the computer runs a Qt software system on the basis of a Qt Creator framework, and the pattern processing module is used for controlling the five-axis galvanometer processing system to process patterns; the error measurement module measures light beam five-dimensional error data; and the error correction module fits an error function through the error data and feeds back the error function to a processing system for correction. According to the invention, pure machine vision five-dimensional error measurement is realized only by using the correction pattern on a single workpiece.

Aspects of the present disclosure provide a method, apparatus, and storage medium including processing circuitry for video decoding. A processing circuit decodes encoding information of a transform block (TB) from an encoded video bitstream. The encoding information indicates one of intra prediction mode information of the TB, a size of the TB, and a main transform type of the TB, and the intra prediction mode information indicates an intra prediction mode of the TB. Processing circuitry determines a context for entropy decoding a secondary transform index based on one of intra prediction mode information of the TB, a size of the TB, and a primary transform type of the TB. The secondary transform index indicates a secondary transform to be performed on the TB in a set of secondary transforms. Processing circuitry entropy decodes the secondary transform index based on the context and performs the secondary transform.

A control method for controlling, by a processing unit (101), a mobile apparatus (200) capable of executing a first mode for performing work and being movable, wherein the mobile apparatus (200) is capable of executing a second mode for performing an operation to respond to the appearance of a subject (306) satisfying a predetermined condition, and the processing unit (101) switches the operation mode of the mobile apparatus (200) from the first mode to the second mode when the processing unit (101) has detected the subject (306).

The invention provides a motion mode processing method and wearable equipment. The method is applied to the wearable device and comprises the steps that after a first instruction is received, whether the wearable device loads map data of a first sports field or not is judged, the first instruction is used for instructing a user to conduct target sports, and the first sports field is the sports field where the user conducts the target sports; after it is determined that the wearable device loads the map data of the first sports field, motion data and positioning data are obtained, the motion data are used for representing motion features when the user conducts target motion, and the positioning data are used for representing motion positions when the user conducts target motion and the position where the wearable device is located; and based on the exercise data, the positioning data and the map data of the first exercise field, displaying statistical data of the target exercise performed by the user. Therefore, the user can be guided to perform redisk analysis of the target exercise, and the user can know own exercise habits, exercise conditions, own stability and the like.

This application provides a method, system, and mask for processing mask patterns. The method is applied to a mask formed by splicing at least two sub-masks, the mask including a splicing region formed based on the overlapping areas of adjacent sub-masks. The method includes: acquiring a mask pattern on the mask and splitting it into a splicing pattern and a non-splicing pattern; predicting the photolithography behavior of the mask pattern using a full-pattern optical proximity effect model of the corresponding mask pattern to obtain a predicted photolithography pattern; when a defect is detected in the predicted splicing photolithography pattern within the predicted photolithography pattern, adjusting the critical dimensions at the position corresponding to the defect in the splicing pattern according to the defect detection result and a critical dimension adjustment rule of the splicing pattern; and outputting the processed target mask pattern. This application improves the efficiency of defect processing for mask patterns located within the splicing region before mask fabrication.

This application discloses a model construction method, apparatus, device, and medium based on federated learning, relating to the field of federated learning technology and applied to cloud data center aggregation nodes, including: obtaining encrypted gradients obtained by participants from encrypting target gradients, and storing the encrypted gradients in a parameter queue; the target gradients are obtained by the participants from training the current model; selecting no more than a target number of encrypted gradients from the parameter queue in a pipeline parallel mode, performing gradient aggregation to obtain an aggregated gradient; decrypting the aggregated gradients in a trusted execution environment to obtain decrypted gradients; using the decrypted gradients to update the current model parameters; obtaining several updated models after the pipeline parallel mode ends; aggregating all updated model parameters to obtain an aggregated model, sending the aggregated model as the current model to the participants, and jumping to the step of obtaining encrypted gradients until the latest model converges. Processing the aggregated decryption and update process in a pipeline parallel mode saves time and improves model training efficiency.

The present invention discloses a protocol reverse parsing method based on multi-feature extraction and semantic simplification, which classifies and extracts features in protocol messages and divides them into fixed features, changing features and associated features. On the basis of extracting these three types of features, a targeted semantic simplification strategy is further introduced. For fixed features, by merging similar semantic fields, duplicate information is reduced and the simplicity of feature expression is optimized. For changing features, a feature pattern processing strategy is adopted to uniformly summarize the changing rules of increasing and decreasing features, and separate the fixed part and the changing part of fixed-length optional features to ensure the integrity of the changing information while reducing dimensions and redundancy. For associated features, the logical relationship analysis between fields is used to dig out hidden field associations, thereby improving the expressiveness and accuracy of the feature library. Through this series of semantic simplification processes, the efficiency of feature extraction is significantly improved.

An apparatus of a base station (BS), the apparatus comprising one or more processors coupled to memory and configured to decode, at the BS, a UE capability message received from a UE, the UE capability message comprising: one or more radio resource management (RRM) artificial intelligence (AI) model IDs of RRM AI models supported by the UE. The RRM AI models are configured to map to transmit beam patterns for a cell for RRM. The processors and the memory can encode, at the BS, assistance information for the UE from a serving cell of the BS, wherein the assistance information comprises: the RRM AI model IDs of the RRM AI models for the BS; and SSB index beam mappings for the RRM AI models by the BS. The UE can be performed RRM using the RRM AI models for the BS with the SSB index beam mappings.

The invention discloses an interface proxy method and device supporting dynamic parameter transmission and a medium, and relates to the technical field of data services, and the method comprises the steps: carrying out the data tracking of a data flow direction, obtaining a data transmission link, and determining a proxy interface based on the data transmission link; receiving the dynamic call request parameter through the proxy interface, and performing dynamic security processing on the dynamic call request parameter to obtain a signature request; based on the signature request, performing polling paging calling to obtain complete calling data; performing paging state real-time storage on the paging process of paging parameter adjustment to obtain a request recovery backup; and performing incremental mode processing on the paging request result to obtain difference merged data. By means of the method, the technical problems that in an existing API proxy method, the data security is poor, and the paging obtaining efficiency is low are solved.

The application relates to a semiconductor device and a manufacturing method thereof, comprising the following steps: providing a substrate, wherein an oxide layer is formed on the substrate; forming a laminated sacrificial layer and a patterned photoresist layer on the oxide layer, and performing a pattern processing on the sacrificial layer to expose part of the oxide layer; performing a first ion implantation process to implant ions into the substrate under the exposed oxide layer; removing the photoresist layer; performing a second ion implantation process to form a well region in the substrate under the exposed oxide layer; and removing the sacrificial layer; wherein the beam intensity of the second ion implantation process is greater than that of the first ion implantation process, and the implantation energy of the second ion implantation process is less than that of the first ion implantation process. The application reduces ion reflection in the ion implantation process, thereby reducing or even avoiding damage to the oxide layer, and effectively improving the performance and stability of the semiconductor device.