48 results about "Numbering system" patented technology

The invention provides a tensor core system for a large language model and a hardware chip, and relates to the technical field of artificial intelligence chips, and the tensor core system comprises an activation matrix processing part which is used for caching and preprocessing activation matrix data read from the interior of a memory to obtain a target activation matrix; the weight matrix processing component is used for caching and preprocessing the weight matrix data read from the inside of the memory to obtain a target weight matrix; the tensor kernel operation component is used for performing matrix floating-point number multiply-add operation on the target activation matrix and the target weight matrix to obtain a floating-point number multiply-add operation result, and accumulating the accumulation matrix and the floating-point number multiply-add operation result based on a preset floating-point number system to obtain a target result matrix; and the accumulation matrix and result processing component is used for performing floating-point number system conversion processing on the target result matrix to obtain a to-be-written result matrix, and writing the to-be-written result matrix into the memory. The method is more beneficial to improving the computing power utilization rate of the chip.

The invention relates to the technical field of spectrum detection, in particular to a Raman spectrum-based rice oil detection method, which comprises the following steps: extracting main and auxiliary peaks to construct a peak group structure, performing sliding comparison to output a conformity region, identifying jump to remove an abnormal signal, calibrating a peak position to generate a correction structure, and performing homing numbering to complete a detection scheme. According to the method, the feature matching rule is constructed by combining the peak position spacing, the intensity ratio and the wavenumber index, the recognition precision of the multi-peak structure in the spectrum is enhanced, non-continuous response signals are recognized and eliminated through derivative change trend and proportion deviation constraint, and the accuracy of stable response extraction is improved. Dynamic calibration and position adjustment of peak positions are realized by utilizing an offset trend sequence, the consistency of a peak group structure in an overall direction and the continuity of a local structure are ensured, a numbering system and structure attribution division is completed after homing sorting, and the integrity of feature expression in a complex sample and the orderliness and reliability of spectral data processing are improved.

The invention provides a building standardization knowledge base building method and system based on a large language model, and the method comprises the steps: carrying out the semantic partitioning of a building specification document set, and obtaining knowledge unit blocks and the logic association relation thereof; processing the knowledge unit blocks by using a large language model, generating a complete coding identifier of an independent knowledge unit and a boundary association identifier of an associated knowledge unit, and determining an association type mark; for the associated knowledge units, determining adjacent knowledge unit blocks and candidate knowledge units according to association type marks and logic association relationships, matching target knowledge units through global positioning identifiers, and allocating same integration numbers; counting the number of independent knowledge units and the number of integrated numbers, and calculating the sum to determine the total quantity of knowledge entities. According to the method, accurate association and integration of knowledge units are realized through a hierarchical numbering system and semantic association strength. According to the method, the construction efficiency of the building standardized knowledge base and the semantic consistency of the knowledge entities can be improved.

The invention discloses a connected domain division method supporting multiple connectivity based on an LBM discrete velocity model, and relates to the technical field of computer-aided engineering, and the method comprises the steps: marking all grid points as full connection points or non-full connection points; performing connected domain numbering on the full connection points and the non-full connection points through a breadth-first search method, storing numbered lattice points and corresponding connected domain numbers to a lattice point connected domain number recording system, and storing a merging relationship among different connected domains to a connected domain numbering system; merging the connected domain numbers in the connected domain number system to obtain a target connected domain number; and updating the connected domain numbers of the grid points in the grid point connected domain number recording system according to the target connected domain numbers to complete connected domain division. According to the method, the connected domains can be efficiently and robustly divided in the three-dimensional geometric models with different qualities, and the requirements of complex engineering scenes are met.

The invention provides a tensor kernel operation component, and relates to the technical field of artificial intelligence chips, and the tensor kernel operation component comprises a decompression layer which is used for carrying out decompression processing on floating-point number elements in a first matrix of a preset floating-point number system to obtain first floating-point number elements in a preset bit format, or carrying out decompression processing on the first floating-point number elements in the first matrix of the preset floating-point number system to obtain second floating-point number elements in a preset bit format; the decompression module is used for decompressing floating-point number elements in a second matrix of a preset floating-point number system to obtain second floating-point number elements of a preset bit format; the calculation layer is used for multiplying and adding all first floating-point number elements corresponding to the first matrix and all second floating-point number elements corresponding to the second matrix to obtain a floating-point number in a target bit format; and the compression layer is used for receiving the floating-point number in the target bit format and carrying out compression conversion processing on the floating-point number in the target bit format to obtain a floating-point number operation result of a number system of the target floating-point number. The tensor kernel operation component provided by the invention is more in line with the matrix multiplication calculation dimension requirement of a large language model.

The invention discloses a multi-role branch chain node state linkage management method and system, and belongs to the field of node collaborative management, and the management method comprises the following specific steps: S101, setting role nodes, adding attribute information of each role node, setting a node basic identification rule, and generating a branch chain according to an intervention approach; the method ensures that different role node numbering systems are independent and not confused, fundamentally avoids the problems of serial number repetition, vacancy or logic fracture, avoids branch structure chaos after role node upgrading, ensures that the hierarchical relationship of nodes in branches before and after conversion is coherent, avoids misadjustment of irrelevant role nodes during exception handling, and reduces system resource consumption. The state adjustment efficiency is improved, manual operation and maintenance links are reduced, manual operation errors are avoided, and meanwhile the response speed of node management and state adjustment is improved.

The invention relates to the technical field of industrial automation control, in particular to an automatic equipment numbering system and method based on an industrial field bus. The method comprises the following steps: broadcasting an initial address seed through a main controller, and sequentially setting incremental addresses for equipment by adopting a chain transmission mechanism in linear topology; exclusive address fields are dynamically divided in the bifurcation topology, buffering is reserved, and hierarchical chain type numbering is started by branch head equipment; monitoring an equipment state in real time, recovering an offline address to an address pool for priority allocation of new equipment, and automatically expanding an upper limit of an address field when the address is exhausted; a conflict risk prediction model based on multi-modal data fusion is introduced, and the crossing from passive processing to active early warning is realized through adaptive adjustment of detection frequency and a preventive reset mechanism. According to the invention, plug-and-play deployment of equipment without manual intervention is realized, and the automatic management level and reliability of an industrial bus network are remarkably improved.

The invention discloses a Chinese character font generation method and system based on four-corner coding, and relates to the technical field of Chinese character generation, and the method comprises the steps: combining four-corner coding with a generative adversarial network and a four-corner consistency loss function, and generating a Chinese character font through a training set composed of a source font and a target font. The method can better adapt to various fonts, and the accuracy of the generated structure is improved; by introducing the four-corner coding information based on the traditional four-corner number system, a generator is helped to better understand the structure details of the fonts, so that the generated fonts have high accuracy in the aspects of the overall structure and strokes, and the generation quality of the fonts is further optimized.

The first aspect of the present application discloses a ship longitudinal bending strength checking method based on component classification and data linkage. The second aspect of the present application discloses a ship longitudinal bending strength checking system for implementing the above ship longitudinal bending strength checking method. The core of the present application is to build an efficient and accurate checking system by scientific component classification and close data linkage, and to deeply couple the professional ship strength checking logic and the data processing capability of the general electronic spreadsheet. The system is based on a unique component numbering system divided according to the plate rack position, combined with the pre-existing double classification of components, to realize the fine management of the ship longitudinal bending strength checking process. At the same time, the system developed based on the Excel platform solidifies the complex checking process into an orderly associated calculation table, ensuring the automatic transmission and efficient processing of data.

The application relates to the technical field of big data processing, in particular to a ship navigation pollutant emission real-time monitoring system, which comprises a multi-source heterogeneous data access unit, a semantic analysis and standardization mapping unit, a flow data time sequence alignment unit, a multi-source collaborative verification unit and a data exchange interface. The system constructs an extensible semantic label library based on the IMO Compendium data numbering system, realizes automatic analysis and standardization conversion of heterogeneous data in combination with an incrementally adapted adaptive mapping rule library, maps multi-source data to a unified time axis through sampling frequency normalization and transmission delay compensation, realizes automatic cross verification and credible fusion of multi-source data by using a shunt state machine of joint verification of numerical consistency, time window alignment and physical consistency, and simultaneously outputs a standardized semantic format and an original data format by adopting a multi-mode sending mechanism. The application reduces multi-source data access cost, improves real-time performance, accuracy and credibility of emission monitoring.

The invention provides a part list automatic checking method based on a configuration BOM, and the method comprises the following steps: S1, constructing a basic number database according to a part function structure, and forming a standard part number according to the basic number database; a plurality of function modules are divided according to the function structure, basic numbers are given to the function modules respectively, the infixes of the standard part numbers are the basic numbers of the function modules to which the parts belong, and the parts of the same function module use the same infixes; in the step S2, variable condition information is set for the parts in the BOM room according to the part configuration information; in the step S3, required part information is input during verification; and S4, according to the required part information, the BOM room information and the vehicle type configuration information, when it is judged that the required part exists in the BOM room, a usage and usage table of the required part on the corresponding vehicle type is output. By establishing a standardized part number system and combining an automatic checking process, high-efficiency verification of the accuracy of the part list can be realized.

The invention discloses a transportation junction route recommendation system based on a large model, and relates to the technical field of route recommendation. The ciphertext coding unit determines all ciphertext codes of this time for response digits generated at this time on the basis of the traffic route recommendation platform every time the service reference data is received; in the process of determining the ciphertext coding, recursive iteration is carried out on the two basic units 0 and 1 in the binary digital system for several times to obtain the basic ciphertext coding used for carrying out the ciphertext coding, and a fixed key does not need to be distributed to a user; according to the method, the burden of frequent key verification during short-term high-frequency use in traditional key encryption is avoided, the scene adaptation defect of key management is solved fundamentally, it is ensured that privacy data such as the real-time position and travel preference of a user can only be decrypted through an exclusive dynamic rule, and privacy safety is remarkably improved.

According to various aspects, a computing device (100) is described herein. The computing device (100) is configured to determine a representation of a first number and a second number in an arithmetic number system, wherein the arithmetic number system includes a positional base system in which each position is associated with a number weight and a digit to represent a number in the arithmetic number system, wherein the number weight is defined by a base of the positional base system and wherein each digit is associated with a set of unique digit elements, wherein the set of unique digit elements is at least two greater than the base, and to perform one or more mathematical operations based on the first number and the second number in the arithmetic number system and to determine a computational result of the one or more mathematical operations in the arithmetic number system.

The application relates to the technical field of form filling and number reporting systems, in particular to a form filling and number reporting system and method based on multi-modal intelligent fusion, equipment and a medium. The application obtains multi-source heterogeneous data through a data acquisition module, establishes a standardized field mapping rule by using a semantic analysis module, analyzes user filling behavior to generate a dynamic data correlation rule in combination with an intention recognition module, an automatic calculation is performed based on the correlation rule by a calculation engine module, hierarchical encryption storage and operation log recording are realized by a secure storage module, and a collaborative distribution module dynamically allocates tasks according to task progress and dependency relationships. The application realizes unified processing of heterogeneous data, improves system adaptability through dynamic data correlation and intelligent task allocation, and effectively improves the efficiency and reliability of form filling and number reporting.

The invention provides a system conversion method and device based on a dynamic mapping rule, and belongs to the technical field of computer systems. According to the system conversion method based on the dynamic mapping rule, the binary data set is dynamically converted into the ternary data sets in parallel by adopting the preset compression code table and the bit parallel conversion logic, so that the storage cost and hardware resource occupation are remarkably reduced, and the storage efficiency is improved. The efficient, flexible and reliable large-scale binary-to-ternary data conversion is realized, the conversion time is greatly shortened, and the conversion throughput is improved.

Proposed are a method, a computer and a computer network that utilize a method of representing and processing numbers in a computer, in which all information transmitted over a system bus to and from a processor is represented in the form of multi-bit numbers in a positional numbering system, wherein all the numbers are represented in a base-b numbering system, each digit in each bit corresponds to its own optical metastable monochromatic laser beam having a wavelength ʎ selected from a range of from ʎ 0 to ʎ b -1, the optical signal of said beam, prior to reaching the processor, is subjected to automatic wavelength selection and sent accordingly over its own optical channel with a given number already in a decimal numbering system to a converter which generates at its outlet a logical 1 in the presence of a signal, or a logical 0 in the absence of a signal, after which this signal arrives at an encoder which converts the given decimal number of the signal into an equivalent number in a binary numbering system, said number together with another similarly processed number arrives at an ALU of the processor and then a binary final number arrives at a decoder where it is converted into an equivalent number in a one-hot base-b code, an active decoder output representing a digit in a bit is converted into an optical laser signal of the corresponding level and said optical laser signal is fed to the system bus. The proposed solutions make it possible to decrease the number of bits in (data, address and command) buses and thus reduce the number of data processing operations and therefore also the machine time by increasing the information content of each bit.

The invention relates to a document processing method and device, electronic equipment and a storage medium. The method comprises the steps that number areas in a to-be-processed document and content hierarchies corresponding to the number areas are recognized; wherein the numbering area is a continuous character interval containing numbering characters or unnumbering titles in the document, and the type of the content hierarchy at least comprises one of a whole-segment title, a first-segment title, a middle-segment title and a text; according to the content hierarchy corresponding to the numbering area and the numbering attribute of the numbering area, determining the numbering level corresponding to each numbering area, and constructing a numbering hierarchy system of the to-be-processed document; and setting a target numbering style for each numbering area according to the numbering hierarchical system. Therefore, the problem that in the prior art, serial numbers in segments are neglected, logic levels are confused, and consequently a serial number system is incomplete and disordered can be effectively solved, and accurate pattern setting of various serial number areas is achieved.

This invention provides an antibody that binds to and stabilizes the trigger receptor 2 (TREM2) protein expressed in human myeloid cells. [Solution] The present invention provides a nucleic acid molecule or a set of nucleic acid molecules encoding an anti-TREM2 antibody or its antigen-binding fragment, comprising three heavy chain complementarity-determining regions (HCDR1, HCDR2, and HCDR3) and three light chain complementarity-determining regions (LCDR1, LCDR2, and LCDR3) containing specific amino acid sequences, as defined by the Kabat numbering system, Chothia numbering system, IMGT numbering system, and combined numbering system. Furthermore, the present invention provides a vector containing the nucleic acid molecule, cells containing the vector, and a method for producing an anti-TREM2 antibody or its antigen-binding fragment, comprising culturing the cells.

The invention belongs to the technical field of underground parking, and particularly relates to an underground parking lot navigation coordinate setting and coordinate number setting method. The method comprises the following steps: firstly, determining a basic center point of an underground parking lot, roughly dividing the parking lot into five areas, namely east, south, west and north, and naming entrance and exit channels of each area; then, according to the positions and directions of the channels, naming rules are formulated for the main channel and the branch channels respectively, and it is ensured that each channel has a unique identifier; navigation coordinate points are set on the main channel and the branch channels according to a certain distance, and are marked by adopting a naming rule containing orientations, serial numbers and serial numbers, so that a navigator can quickly position the navigation coordinate points. In addition, the scheme further provides setting suggestions of the intersection points, and the numbering mode can be flexibly selected according to the layout of the basement. According to the scheme, the navigation efficiency and the user experience of the underground parking lot are effectively improved through a clear naming rule and a numbering system.

1. Name of the design product: Graphical user interface of the queuing system for electronic devices. 2. Use of the design product: For an electronic device. 3. Design points of the design product: Lies in the graphical user interface. 4. Picture or photograph that best shows the design points: Front view. 5. The electronic device is of conventional design, and other views of the electronic device are omitted. 6. Use of the graphical user interface: For queuing and calling numbers and viewing information related to the currently called person and the current queuing situation. For example, it can be used for queuing and calling numbers in the Putonghua Proficiency Test. 7. Human-computer interaction method of the graphical user interface: The initial interface of the calling number system in the front view. If the calling number starts, the interface jumps to the interface in the queuing and calling number as shown in the interface change state diagram; "XXX" in the above interface represents replaceable text and / or numbers, and the gray rectangular color block in the upper left corner of the interface represents replaceable content pictures.

The invention discloses an internet-based enterprise management function teaching platform, which relates to the technical field of function teaching, and is characterized in that a secure transmission module is arranged to convert learning data, a mapping table is constructed immediately before conversion, mapping conversion is carried out on the learning data according to the composition of characters, and the mapping table is deleted after conversion; in this way, key information traces in the data conversion process are thoroughly eliminated, the risk of information leakage is effectively prevented, a mapping set corresponding to each group of standard combination is obtained based on a plurality of groups of number system combinations obtained through learning data and the standard combinations, and the mapping set comprises a plurality of mapping word strings. The mapping word string is obtained and selected by the characters filled in the standard, and the combined mapping sequence of the number system combination is randomly selected from the corresponding mapping set, so that the collision resistance and complexity of the mapping conversion result are enhanced, and the safety of the learning data in the transmission process is further ensured.

The application discloses a method for reducing the loss of large dynamic FFT operation by truncation and number system conversion, and relates to the field of radar signal processing. Firstly, by selecting appropriate rotation factor quantization bits and complex multiplication truncation bits, the full frequency domain precision loss is ensured to be not more than a fixed value under the condition of small signal input; then, whether to retain the bit width increased by complex addition operation is judged according to the order corresponding to the number of FFT points required to be operated; finally, according to the operation bit width truncation rule, after completing the entire FFT operation, the FFT operation result represented by fixed-point numbers is converted into the FFT operation result represented by floating-point numbers. The application can effectively retain the spectrum information under the condition of small signal input, avoid the signal distortion caused by the truncation in the FFT processing process, greatly improve the dynamic range, and provide good data support for subsequent fine processing.

The invention relates to the technical field of unmanned aerial vehicle airborne platforms, in particular to a modularized unmanned aerial vehicle airborne multi-sensor integrated platform, which comprises the steps of collecting equipment names and installation numbers, extracting sensing directions and reaction time ranges, establishing serial numbers, comparing time starting and ending, extracting association sequences, mapping equipment path position numbers and rearranging node sequences. And combining perception and energy numbers, generating a reorganizing table, binding the reorganizing table to track nodes, and outputting a regulation and control structure. The method comprises the following steps: acquiring equipment names, installation numbers, sensing directions and reaction time parameters, constructing a serial number system, comparing and extracting an equipment combination sequence based on a coverage relation of starting and ending time, combining position numbers of an image focuser and a signal switcher, carrying out node rearrangement on an equipment operation sequence, and carrying out node rearrangement on the equipment operation sequence. And according to the perception area and the energy number updating rule, generating an activation sequence reorganizing table, binding the device number to the track node, and outputting a signal calling and number mapping result.

Systems and techniques may generally be used for controlling a low earth orbit (LEO) satellite. An example technique may include sending a control signal to a thruster of the LEO to maintain a trajectory of the LEO satellite along a particular orbit, and capturing, at a sensor device, at least one photon. The example technique may include generating, for example using processing circuitry at the LEO satellite, a random number based on the at least one photon.

The invention relates to the field of edge intelligent control, and discloses an intelligent central control system based on ARM gateway equipment, which comprises a plurality of cameras, acquisition entrance, hall and counter camera data and a plurality of pieces of ARM gateway equipment. The entrance ARM gateway device generates entrance face information, and the hall ARM gateway device generates hall face information. The counter ARM gateway equipment receives and matches the face information, and generates a preprocessed information sequence; determining a customer retention probability corresponding to each piece of preprocessing information in the sequence, and generating a priority processing sequence; generating target customer information containing a target employee identifier and a predicted business category according to the priority processing sequence, the scheduling information and customer historical business information; and obtaining a corresponding actual service category according to the target customer identifier, comparing the actual service category with the predicted service category, generating target number calling information, and pushing the information to a number calling system. Therefore, the customer service satisfaction is improved, and the customer retention probability is improved.

The invention discloses a user behavior data acquisition method for a fission marketing platform, and relates to the technical field of data acquisition. According to the method, a task linkage coding field is established through a task node numbering system and an inter-task association rule matrix, so that a user behavior state has path identification and jump tracking capabilities; a user relation network diagram is further combined, a user behavior collaborative index table is constructed, a behavior collaborative collection triggering mechanism based on the fission relation among multiple users is achieved, and the limitation of traditional single-point triggering collection is broken through; and a task chain type acquisition path diagram is constructed by identifying a jump boundary of a key path in a task chain and splicing data segments, so that the continuity, the integrity and the semantic consistency of acquisition are effectively improved.

An embodiment of this specification discloses a method, device, and electronic device for detecting abnormal content. The method for detecting abnormal content includes: obtaining numerical feature data of a target user, and converting the feature data into data in a preset number system; determining a representation matrix corresponding to the converted feature data based on preset representation vectors corresponding to the digits of each data bit of the preset number system in a preset representation matrix; using a pre-trained encoding network to perform encoding processing on the representation matrix corresponding to the feature data to obtain a representation vector corresponding to the feature data; and inputting the representation vector corresponding to the feature data into a detection model for detecting abnormal content to obtain a detection result of abnormal content for the target user.

The invention relates to the field of BIM modeling design, in particular to a drawing method of a support hanger, computer equipment and a storage medium. The method comprises the following steps: acquiring a support model and a construction specialty of each support hanger in a global modeling environment; determining a first serial number of each support hanger in the global modeling environment according to the support model and the construction specialty of the support hanger, and generating a support structure drawing of the support hanger corresponding to each first serial number; determining a plurality of target supports and hangers in each construction project in response to a project numbering request of a user; determining a second number of each target support hanger according to the support model and / or the construction specialty of the target support hanger; performing duplicate removal screening on the first numbers corresponding to the plurality of target supports and hangers, determining a plurality of target first numbers, and exporting a support structure drawing corresponding to each target first number; and generating and exporting a corresponding project overview drawing based on the second number. While the drawing is lightly output, the effectiveness of the drawing on construction guidance is improved.

The invention relates to the technical field of laser marking, and discloses a laser marking method and system based on work order SN number skip management, and the method comprises the steps: obtaining a work order data structure which comprises the total number of PCBs, the number of jointed boards, a prefix character string, a suffix character string, a system type and a start number; calculating a current serial number character string, and filling the current serial number character string into the variable object in the marking template file to obtain marking content data; correcting a first marking coordinate corresponding to the marking content data to obtain a second marking coordinate; and the laser head is controlled to move to the second marking coordinate and execute laser marking operation, the current serial number character string and marking result data are uploaded to the MES system after marking is completed, the efficiency loss caused by frequent template programming and modification is avoided, and the production management automation level and the data traceability in the intelligent manufacturing environment are improved.