40 results about "Computation type" patented technology

The embodiment of the invention provides a computing resource distributing method, a distributed type computing method and a distributed type computing device which are applied to a distributed type system. The computing resource distributing method comprises the following steps: grouping computing resources of computing nodes of the distributed system according to a computing type; establishing a computing tool corresponding to the computing type, and distributing the computing resources in the computing nodes to the corresponding pool according to the computing type; and while performing computing task processing, firstly searching the computing pool corresponding to the computing type of the computing task, processing the computing task of the computing type through computing resources of the computing pool, so that the computing resources of the same computing type only process the computing task of the computing type. The computing resource distributing method is convenient for managing the computing resources, so that the computing resources are effectively utilized.

The invention provides a big data platform system capable of realizing separation of computation type applications and data, and a running method therefor, and belongs to the technical field of cloud computing and big data. The system can realize decoupling and separation of the applications and the data, and solve the problems of difficulty in migration from applications to data and multi-table associative computing of the data due to a strong dependence relationship between the applications and the data in an existing big data platform technology. The running method for the big data platform system comprises: virtually combining hard disks of a plurality of physical machines into a virtual hard disk; dividing the virtual hard disk into a plurality of storage volumes; and receiving a request and creating a virtual machine (which can be just all of a CPU and a memory of a complete machine, but still is the virtual machine) used for running the request, wherein the storage volumes are mounted for the virtual machine according to the demands of the request.

A broadcast group managing router (1) transmits a broadcast packet by attaching thereto a calculation type address which allows to arithmetically obtain a delivery destination by calculation, and each calculation type address calculating router (2 to 4) calculates the calculation type address so as to determine the delivery destination, to thereby conduct the broadcast communication.

The invention discloses a transform operation method, a transform operation device, a coordinate rotation digital computation method and a coordinate rotation digital computation device. A transform operation approximate evaluation method comprises the following steps: judging a type identifier of a function to be resolved, and searching for an iteration parameter corresponding to the type identifier; and according to the iteration parameter and the initial value of the function to be resolved, performing the iterative operation of the function to be resolved. In the invention, the corresponding iteration parameter is selected according to the preset type identifier; and when the methods and devices are used for resolving functions of different types, the configured and preset parameter of software are regulated, but the structure of hardware is not changed; one set of implementing resource is used to realize various computation types; and the implementation method is simple and normalized.

In order to make it possible for an operator of a machine tool to easily calculate machining conditions and prevent calculation errors and input errors by the operator, this control device for a machine tool, which machines a work piece on the basis of a machining program, is provided with: a database that stores calculation formulas corresponding to types of calculation, and types of parameters required to execute said calculation formulas; an input unit that selects a type of calculation; and a display unit that displays, on the basis of the type of calculation selected by the input unit, the parameter to be input.

Disclosed are an apparatus and method for controlling transmission power of a terminal in a wireless environment. The apparatus for controlling transmission power of a terminal includes an optimal transmission power calculation unit configured to calculate an optimal transmission power value for a terminal of a type, the terminal being accessible to one or more wireless networks; a transmission power setting unit configured to, when an access of the terminal to at least one of the one or more wireless networks is sensed, provide the terminal with optimal transmission power values corresponding to the terminal and the network accessed by the terminal.

The invention discloses an activation function hybrid calculation method and system based on CORDIC, and belongs to the technical field of function calculation. The method comprises: a core control module processing a to-be-evaluated x according to a calculation type t to obtain an input value a and inputting the input value a into a hyperbolic rotation mode CORDIC module; performing operation according to the input value a to obtain xn and yn, adding the xn and yn through an adder to obtain an input value b, inputting the input value b into a linear vector mode CORDIC module for division operation to obtain z'n, and processing and outputting the z'n according to the calculation type t. The system comprises a core control module, a hyperbolic rotation mode CORDIC module and a linear vectormode CORDIC module, wherein the hyperbolic rotation mode CORDIC module and the linear vector mode CORDIC module are respectively connected with the core control module. The method aims to overcome the defects that in the prior art, the sigmoid function and the tanh function are low in calculation precision and the calculation range is not easy to expand, the calculation precision of the sigmoid function and the tanh function can be improved, and the data range can be expanded.

The invention provides a fusion calculation method and a readable storage medium. The fusion calculation method comprises the steps of acquiring a micro-control instruction sequence and distributing to multiple started threads, and the micro-control instruction sequence comprises the calculation type of a calculation task; determining a fusion calculation mode based on the calculation type of thecalculation task; and scheduling the thread to execute the calculation task according to the fusion calculation mode.

The invention discloses DPD (Digital Pre Distortion) self-adapting method and a device based on symmetric multiprocessors, wherein the method comprises the following steps of: carrying out correlation decoupling upon specific processing modules which meet preset decoupling conditions during a DPD self-adapting progress, obtaining a correlation decoupling result, wherein the decoupling conditions comprise a computation amount, a computation type, decoupling possibility, decoupling complexity, a decoupling risk and a decoupling efficiency speed-up ratio; carrying out parallel processing upon uncorrelated data which meets a preset correlation degree in the correlation decoupling result, obtaining parallel processing data; processing the parallel processing data by result merging, obtaining result merging data. By adopting the method and the device disclosed by the invention, load of different cores of the processors can be balanced, and synchronously, an execution speed can be improved effectively, the DPD self-adapting progress can be accelerated and convergence can be achieved, as a result, a non-linear change of a system can be preferably tracked, and efficiency and a performance of the system can be enhanced.

The invention belongs to the technical field of knowledge graph representation learning and ontology construction, concretely relates to an ontology inclusion axiom learning method based on linear programming. The method comprises the following steps: constructing a representation learning model SetE, inputting entity data, type data and relationship data of a knowledge graph Abox into the representation learning model SetE, calculating inclusion relationships among the type data, and extracting the inclusion relationships to form axioms. According to the method, predicate features can be captured by using a linear representation learning model, the inclusion relation is concluded and reduced into linear programming to obtain the ontology inclusion relation, and the corresponding ontologycan still be efficiently learned under the condition that the knowledge graph is incomplete.

The invention belongs to the technical field of mathematical calculation, and discloses a calculation method and device for function derivation, calculation equipment and a storage medium. The methodcomprises the steps: enabling a calculator to obtain function information corresponding to a to-be-calculated function expression input by a user; extracting a target variable type contained in the function information, and determining a target equation type corresponding to the to-be-calculated functional expression according to the target variable type; searching a corresponding derivation calculation model in a preset model database according to the target equation type; constructing an auxiliary function according to the to-be-calculated functional expression, and searching a partial derivative result corresponding to the auxiliary function in a preset partial derivative database; and substituting the partial derivative result into the derivation calculation model, and obtaining a model output result. By means of the mode, various types of functions can be derived quickly and accurately, and the technical problems that in the prior art, derivation calculation is single in type, time is consumed in calculation, and accuracy is hard to guarantee are solved.

One or more embodiments of the present specification provide a method for scheduling calculation service for business process contracts. The method comprises the steps of running scheduling processes independent of the nodes on the node devices corresponding to at least part of the nodes, wherein the scheduling processes are used for scheduling the computing service processes independent of the nodes and running on the node devices. And for each calculation task in the business process contract, according to the calculation type of the calculation task, scheduling an available calculation service process corresponding to the calculation type on the same node equipment as a calculation service process serving the calculation task by the scheduling process corresponding to the node participating in the calculation task. In this way, in the process of executing the business process contract by the block chain network, the calculation service process occupied by each calculation task can be called through the scheduling process to execute the calculation task.

The invention discloses a method and device for achieving ABAQUS calculation tasks. The achieving method comprises the steps of obtaining calculation instructions input by users; analyzing the calculation instructions, and determining calculation types and calculation parameters; enabling the calculation parameters to be substituted in pre-allocated calculation command statements corresponding to the calculation types, and carrying out the calculation command statements. By means of the technical scheme, the achievement process of the ABAQUS calculation tasks does not need to be by means of command lines, the calculation tasks can be finished only by means of simple operation based on a graphic interface, operation difficulty is greatly reduced, execution efficiency is improved, and the error probability is reduced; besides, computing resources can be reasonably distributed and utilized by the analysis of resource request instructions input by the users and the execution of computing resource scheduling statements, and the computing resource utilization rate is improved; in addition, by the judgment of validity of information input by the users, safety and stability in the running of an ABAQUS calculation program can be guaranteed well.

The invention discloses a processing method for a computational formula in a data table. The processing method comprises the following steps of: judging the computation type of the data table; and reading the computational formula for items which correspond to the computation type of the data table for computation. By the processing method for the computational formula in the data table, the items can be automatically selected for the current computation of the data table, and the computation of the data table is more flexible, simpler and more convenient. The invention also provides a processing device for the computational formula in the data table.

The invention provides a data processing method and related equipment. The method comprises the following steps: acquiring a plurality of buried point data; determining a calculation type corresponding to each buried point data, and determining target buried point data according to the calculation type corresponding to the buried point data; or determining the calculation type and a calculation rule corresponding to each buried point data, and determining target buried point data according to the calculation type and the calculation rule corresponding to the buried point data; and processing the burying point data and the target burying point data to generate a business index. In the process, the slave library or the data warehouse does not need to be configured for storing the operation data of the system, and the operation data is subjected to related calculation to obtain the business indexes. As the slave library or the data warehouse does not need to be configured in each step involved in the embodiment of the invention, the operation cost of the system is greatly reduced.

The invention provides a work order management method, device and equipment and a readable storage medium. The method comprises the steps: when an equivalent work order capable of being equivalent to a current work order exists, obtaining the calculation type of the preventive maintenance action, wherein the calculation type is any one of the first calculation type, the second calculation type, the third calculation type and the fourth calculation type; and setting a related work order associated with the equivalent work order according to the calculation type of the preventive maintenance action and the equivalent work order. According to the method and the device, when the current work order is equivalent to the equivalent work order, the related work order associated with the equivalent work order is automatically set according to the calculation type of the preventive maintenance action and the equivalent work order, so that the manual operation is omitted, the work order setting efficiency is improved, and the efficiency of the preventive maintenance action is improved.

The invention discloses a transform operation method, a transform operation device, a coordinate rotation digital computation method and a coordinate rotation digital computation device. A transform operation approximate evaluation method comprises the following steps: judging a type identifier of a function to be resolved, and searching for an iteration parameter corresponding to the type identifier; and according to the iteration parameter and the initial value of the function to be resolved, performing the iterative operation of the function to be resolved. In the invention, the corresponding iteration parameter is selected according to the preset type identifier; and when the methods and devices are used for resolving functions of different types, the configured and preset parameter of software are regulated, but the structure of hardware is not changed; one set of implementing resource is used to realize various computation types; and the implementation method is simple and normalized.

The invention relates to an algorithm variable data intelligent central station system, which comprises: a data docking module configured to manage an existing data source of an internal system, obtain the data source from an external system and generate a derivative variable; the variable development module that comprises a foreground self-development unit and a background code development unit, wherein the foreground code development unit is configured to develop common functions of derivative variables through internal packaging of the system and provide page selection and use, and the background code development unit is configured to develop a new variable in a dragging manner through a visual interface, and derive variable logic capable of running and calculating on an online system; and the variable logic calculation module that is configured to perform real-time calculation according to the configuration and parameters of the derivative variables and dynamically output calculation results of the derivative variables according to the calculation types of the variables. According to the algorithm-based variable data intelligent central station system, the coverage degree of big data types can be improved, the real-time online processing capability is high, and the processing time is short.

The embodiment of the invention provides a data processing method and device and a computer readable storage medium, and the method comprises the steps: receiving an acceleration task from a DSP, theacceleration task carrying a calculation type and calculation data; determining hardware corresponding to the calculation type and a driving function of the hardware, wherein the hardware is a specialchip for operator operation in deep learning; and driving the hardware by the driving function to process the calculation data. According to the embodiment of the invention, the data processing efficiency can be improved.

The invention discloses a distributed computing system and method. The distributed computing system comprises n computing units, a switch and a computing center, the n computing units are in communication connection with the computing center through the switch; and the n calculation units are divided into m groups according to the calculation types, wherein m is the number of the calculation types. The calculation units are grouped according to the calculation types, and the calculation units in the corresponding groups carry out calculation processing on the information of the corresponding calculation types, so that matrix calculation service is provided.

The invention relates to the technical field of communication, and discloses a signaling service subsystem. The signaling service subsystem comprises a processing method acquiring module and a counter value calculating module, wherein the processing method acquiring module is used for acquiring a processing mode of a current counter from pre-set start index positions, recorded in a counter information list, of all types of counters of a counter group and an index position of the current counter, the counters of the same processing modes are the counters of the same type, indexes of each type of counters are continuous, the counter value calculating module is used for carrying out accumulation to a value of the current counter and times of the current counter, and calculating the value of the current counter according to the processing mode. The invention further discloses an expanding counter processing system and method comprising the signaling service subsystem. According to the signaling service subsystem and the expanding counter processing system and method, the counters with the same processing mode are classified as a type, only the start index positions of each type of counters are provided, a maintenance-redundant messy calculating type list of an expanding calculator is avoided, and maintenance is convenient.

The invention belongs to the field of graph analysis, and relates to a graph triangle counting method based on bit operation. A traditional graph triangle counting method often relates to a large number of matrix multiplication operations and complex control logic. The requirements of quick community discovery and the like in a super-large-scale social network graph cannot be met. According to thegraph triangle counting method based on the bit operation, the graph triangle counting problem is converted into a series of simple bit operations such as Boolean AND operation and bit counting, thecalculation type and the control logic are greatly simplified, and therefore the number of triangles in the graph is efficiently and quickly calculated. Meanwhile, the calculation characteristics of the triangle counting method provided by the invention are suitable for being applied to a novel in-memory calculation architecture.

The invention provides a work order management method, device, equipment and readable storage medium. The method includes: when there is an equivalent work order that can be equivalent to the current work order, obtaining the calculation type of the preventive maintenance action, and the calculation type is the first calculation type, the second calculation type, the third calculation type and the fourth calculation type. Any one of the calculation types; according to the calculation type of the preventive maintenance action and the equivalent work order, set a related work order associated with the equivalent work order. Through the present invention, when the current work order is equivalent to the equivalent work order, the related work order associated with the equivalent work order is automatically set according to the calculation type of the preventive maintenance action and the equivalent work order, which saves manual operation, Improved work order setup efficiency, resulting in more efficient preventive maintenance actions.

The invention relates to the field of power systems, in particular to an online assistant decision making method for large power grid restoration. Corresponding steps of the online assistant decision making method are carried out in different selected working modes, and the working modes include a real-time assistant mode and a research and analysis mode. The online assistant decision making method has the advantages that the merit that dispatchers and equipment operators of a power plant and a transformer substation are familiar with the traditional plan type system restoration scheme when the traditional plan type system restoration scheme is adopted is kept in the real-time assistant mode, the work intensity of plan formulation personnel is relieved, the assistant decision making speediness, accuracy and reliability for power system restoration are embodied, and shortcomings that an actual condition of a power grid which is developed quickly is inconsistent with a designed condition of a plan, restoration paths involved in the plan are infeasible and the like are overcome specifically; almost all computation types which are possibly involved in formulating a common power grid black-start scheme are integrated in the research and analysis mode, typical power grid partition computation for large power grid restoration is added into the research and analysis mode, a scheme draft generating function can be realized, and the expenditure of time spent on integrating different result files is greatly saved.

The invention provides a method for a heterogeneous computing system to execute a computing task. The method comprises the following steps of: determining a data volume, a computing type and code complexity which need to be computed according to a current computing task; determining computing equipment with the shortest computing time in the heterogeneous computing system according to the data volume, the computing type and the code complexity; and calculating the current calculation task by adopting the calculation equipment with the shortest calculation time. According to the method, the computing task can be executed on the corresponding equipment with the shortest computing time, so that the advantages of the heterogeneous computing system are fully exerted.

The embodiment of the invention provides a data processing method and a processing core. The method comprises the steps of receiving a calculation instruction, wherein the calculation instruction carries input data and a calculation type, and the calculation type comprises a modular multiplication type or a modular exponentiation type; according to the calculation type, determining a state sequence of a state machine, each state in the state sequence corresponding to a preset operation step; switching the state of the state machine according to the state sequence, outputting a first control signal to a corresponding multiplexer in each state of the state machine so as to control calculation data input into a Montgomery calculation unit through the multiplexer in each state, and executing operation steps corresponding to all the states based on the calculation data by the Montgomery calculation unit, wherein the calculation data comprises input data and / or an intermediate result obtained in a previous state; and after the operation step corresponding to the last state of the state sequence is executed, obtaining a calculation result of the calculation instruction. According to the embodiment of the invention, on the basis of reducing the circuit area overhead, the speed and concurrency of modular operation can be improved.

The invention provides a configuration method and a device, and a computing system, and belongs to the field of embedded computing. Wherein each computing unit of the computing system comprises at least two types of operators; the method comprises the steps of receiving a to-be-configured target algorithm; analyzing the calculation types contained in the target algorithm and the number corresponding to each calculation type; configuring a target operator for each target computing unit according to the computing types contained in the target algorithm and the number corresponding to each computing type; according to the target algorithm and the target operator corresponding to each target computing unit, determining a network structure formed by all the target computing units; and accordingto the target operator corresponding to each target computing unit and a network structure formed by all the target computing units, generating a configuration file of the computing system for realizing a target algorithm. Different types of algorithms are realized through the same computing system, and the parallel processing capability and the hardware resource utilization rate of the computingsystem are improved.