110results about How to "Improve computing performance" patented technology

System(s) and method(s) that facilitate utilizing pre-cached disk space. Pre-cached memory space within a storage device is identified, and a subset of the pre-cached memory space is pre-populated with data so that the data can be selectively and dynamically accessed. During use of a computer (e.g., in a web-browsing session) a subset of the pre-stored data can be dynamically and selectively exposed to the user as a function of user and / or computer application state. Pre-storage of the data on pre-cached memory of the computer mitigates delayed data access (e.g., due to insufficient transmission bandwidth) thereby enhancing user computing experience. The user can utilize the device without having to distinguish between pre-cached and free memory. In other words, the operating system can present the cached memory to the user so that it appears as free memory without the user having to direct the system to do so.

The invention provides a vector and matrix operation-based calculation-supported encryption method, which comprises the following steps: 1. converting initial data into a vector and encrypting the vector by using an encryption matrix so as to obtain an encrypted outsourcing vector; 2. converting a calculation parameter into a vector and encrypting the vector by using the encryption matrix so as to obtain an encrypted calculation parameter vector; 3. carrying out operations, such as scalar product, addition, subtraction and multiplication, on the encrypted outsourcing vector and the encrypted calculation parameter vector so as to obtain an encrypted calculation result; and 4. obtaining the plaintext of the calculation result through carrying out an operation on the encryption matrix and the encrypted calculation result by a user. According to the method provided by the invention, an encryption scheme with IND-COA security is always used for the encryption of data or the operation of encrypted data, thus the function of protecting personal privacy is exerted; and meanwhile, the method has excellent operating performance in the fuzzy retrieval of encrypted character strings, excellent calculation performance for service providers and a moderate storage and communication load.

The invention discloses a super-dense edge computing network mobility management method based on deep reinforcement learning. The super-dense edge computing network mobility management method includesthe steps: establishing a communication time delay model, a computing model, a QoS model and a service cost migration model according to environment information and processing resource information; establishing a mobile management model according to the established model information, simplifying the problem by adopting a dynamic loss queue technology and a Lyapunov optimization method, and abstractly describing a dynamic change process of an ultra-dense edge computing environment of the mobile management model by adopting a discrete time Markov decision process; and establishing an algorithmbased on deep reinforcement learning according to the abstract model and obtaining an optimal mobility management decision. According to the super-dense edge computing network mobility management method, for a super-dense edge computing network, the mobility management decision is small in limitation and good in mobility, and on the premise of considering the integrity, dynamics and balance of thesystem, the optimal decision of the association network and task allocation in the user moving process is realized.

The invention discloses a fixed point vector processor and a vector data access controlling method thereof, relates to a vector processor used for online time series prediction and aims to solve the problems that an existing vector processor incapable of optimizing a specific method is poor in universality and cannot satisfy requirements on online computation. The fixed point vector processor comprises a program counter, a microcode memorizer, a vector memorizer, an arithmetic logic unit and a data control unit. A complete fixed point vector processing procedure is formed by every signal processing procedure of the program counter, the microcode memorizer, the vector memorizer, the arithmetic logic unit and the data control unit. By means of ALU (arithmetic logic unit) designing, an ALU structure of each data path can be changed flexibly according to computation needs, flexible configuration of an instruction set is achieved, and accordingly, the fixed point vector processor and the vector data access controlling method are applicable to occasions required by complex computation.

The invention discloses a circuit breaker assembly quality problem tracing system and method based on large data analysis. The problem that an ancillary system cannot meet the requirement of massive data analysis in a circuit breaker assembly production process is solved. The new idea of quality circuit breaker assembly quality problem tracing is provided from the aspect of data mining. The method comprises the steps that Spark is used to build a large data analysis platform; data preprocessing is carried out in combination with business; K-Means is used to build an assembly quality problem classification model based on distance; an Apriori association rule algorithm is used to build an influential factor analysis model; and assembly quality problem tracing is realized. The assembly quality problem classification model completes the deficiencies of mechanism classification or experience classification from the aspect of data characteristics. According to the invention, the system and method are capable of storing and analyzing massive data, have the advantages of high-performance redundancy, real-time and efficient data analysis and storage, are used for circuit breaker assembly quality problem tracing, and provide suggestions and references for circuit breaker assembly quality production.

The invention relates to detection of epileptiform activity. In order to accomplish a mechanism with improved specificity to epileptiform activity and with the capability to detect specific type of epileptic patterns, brain wave signal data obtained from a subject is decomposed into at least one predetermined subband, each subband being indicative of a specific type of epileptiform activity. The subband-specific output data obtained represents a time series of a quantitative characteristic of the brain wave signal data. At least one measure is determined for any one or more of the at least one predetermined subband, the at least one measure belonging to a measure set comprising a first measure indicative of the entropy of the subband-specific output data and a second measure indicative of a normalized form of k:th order central moment of the subband-specific output data, where k is an integer higher than three. The presence of a specific type of epileptiform activity may be detected based on the at least one measure of the respective subband.

The invention discloses a power distribution network flow computing method based on equivalent node injecting current transformation. The method comprises the steps that firstly based on a traditional fundamental principle of a Newton-Raphson method, the equivalent node injecting current transformation is adopted, a non-linear flow equation which corresponds to PG nodes in a power distribution network is transformed into a linear equation, accordingly a Jacobian matrix in a flow iteration calculation formula can be simplified, most elements of the matrix are enabled to be constants, repeated calculation is not needed, and calculated amount is reduced; then the Jacobian matrix is calculated through a sparse technology, memory space is greatly saved, and calculation speed and efficiency of an algorithm are improved; finally, accuracy and effectiveness of the algorithm can be verified through test cases. In addition, through compassion of calculation performance of different algorithms, the computing method has the advantages of being high in calculation speed, high in calculation efficiency, good in convergence property and strong in looped network processing capacity.

The invention discloses a vehicle-mounted ad hoc network security authentication method based on a Cuckoo filter. The method comprises the following two steps: 1) identity authentication between a vehicle and an infrastructure (V2I), wherein the step comprises the following four specific processes: system initialization, initial handshake, message signature and massage authentication; and 2) identity authentication between vehicles (V2V), wherein the step comprises three specific processes: group key generation, group message signature and group massage authentication. The method is realized based on software without depending on any other specific hardware equipment; an unparalleled linear pairing method is adopted in the authentication phase, and the unparalleled linear pairing authentication method is suitable for both V2I and V2V authentication, so that the method has efficient computing power and lower cost; and in the batch certification phase, the Cuckoo filter and the binary search technique are adopted to detect whether signature is effective, so that efficiency of the authentication phase is improved to a great extent.

The invention relates to a method and a device for increasing a video encoding speed, and belongs to the field of image processing. The method and the device are characterized in that a CPU (central processing unit) of a video compression encoder performs logical operation for infra-frame prediction or inter-frame prediction mode selection, loop filter and entropy encoding, a GPU (graphics processing unit) of the video compression encoder performs numerical operation for movement estimation, movement compensation, transformation and quantization and inverse transformation and inverse quantization, and the device comprises a CPU framework flow line and a GPU framework flow line which are parallel to each other. The method and the device have the advantages that a heterogenous system with a combination of a CPU framework and a GPU framework is comprehensively utilized for increasing video compressing and encoding speeds, so that videos are effectively compressed on the premise that the quality is guaranteed, the encoding elapsed time is greatly shortened, the integral performance of the encoder is greatly improved, the integral complexity of the encoder is assuredly unchanged basically, the instantaneity of the video encoder is greatly improved, and the method and the device can be applied to real-time encoding and decoding places.

The invention discloses a method for calculating the power flow of a radial power distribution network based on variable substitution. The method comprises the following steps of: firstly, converting 2 / 3 of unbalance equations into linear equations in power flow calculation by using variable substitution according to the special topological structure and characteristic of the radial power distribution network, wherein the other 1 / 3 of unbalance equations are also relatively simple; then, because nonzero elements of a jacobian matrix formed by an unbalance equation set are relatively fixed, respectively calculating the invariant elements and a small number of variable elements of the jacobian matrix by using a sparse technique, wherein the invariant elements do not need to be repeatedly calculated, and a calculation formula of the variable elements is also relatively simple; and finally, carrying out iteration solving on the unbalance equation set to obtain the power flow of the system. The calculation speed and efficiency of a power flow algorithm are improved foundationally through the variable substation and the sparse technique; and the method has the characteristics of clear calculation process, easiness for programming and the like. Finally, the effectiveness and efficient calculation capability of the method are proved through a plurality of test cases.

The invention belongs to the field of data processing technologies, and particularly relates to a distributed extreme learning machine optimization integrated framework system. The distributed extreme learning machine optimization integrated framework system comprises a data distributed storage module, a Stack-Autoencoder feature extraction module, a distributed calculation module and a result output module, wherein the data distributed storage module stored data in a distributed manner, and analyzes and determines an ELM hidden layer; the Stack-Autoencoder feature extraction module carries out feature learning on the data, obtains compressed input data, and performs normalization processing and feature extraction on the input data; and the distributed calculation module obtains an overall optimal weight through carrying out mapping and reduction processing on the input data. The distributed extreme learning machine optimization integrated framework system is more accurate in pattern classification of big data, solves the overfitting problem caused by too much nodes of a single-layer ELM, allows the calculation partitioned blocks of a high-dimensional matrix to be conducted in parallel, increases calculation efficiency, and saves memory resources since data does not need to be read into the memory in advance.

The invention provides a weighted DV _ Hop node positioning method based on a minimum mean square error criterion, and the method comprises the steps: setting other anchor nodes to have the same weight under the minimum mean square error criterion, and obtaining a preliminary estimation value of the average distance of each hop of the anchor nodes; setting weights for other anchor nodes for the second time, and obtaining the corrected average hop-by-hop distance of the anchor nodes under the minimum mean square error criterion; calculating the connectivity of each adjacent anchor node and theaverage distance error of each hop; calculating the average hop distance of the unknown node and the estimated distance from the unknown node to the anchor node according to the average hop distance of the adjacent anchor nodes and the normalized weight value; and setting a linear solution of a weighted least square method as an iterative initial value, minimizing a weighted distance estimation error objective function through a nonlinear weighted iterative method, and taking an iterative result as a positioning coordinate of an unknown node. According to the method, the more accurate averagehop distance between the anchor node and the unknown node can be obtained, the calculated amount is small, and the positioning accuracy and stability are improved.

Provided in the invention is a design method for access of a flexible multi-state switch to a distribution network. On the basis of simulation of a multi-scene power flow, distribution network weak positions including distribution network node voltage limiting exceeding and line load heavy loading and the like after access of a distributed power source are calculated and analyzed and a flexible multi-stage switch access location is determined; a system optimization model of access of the flexible multi-stage switch to the distribution network using economic cost minimization as an objective function is established, wherein the objective function includes a yearly conversion cost and a system operation loss cost of initial investment of the flexible multi-stage switch and various constraintconditions are met; with an improved particle swarm optimization method based on a genetic algorithm (GA-PSO), an optimization design model including discrete variables and continuous variables of the flexible multi-state switch is calculated, mixed coding is carried out on the complicated discrete variables and continuous variables, and crossover and mutation operations of the genetic algorithmare introduced into a particle swarm optimization method for improvement; and thus an optimization design plan for access of the flexible multi-state switch to the distribution network is obtained byoptimization calculation.

The invention provides a text processing method and a text processing device, which are used for solving the problem of poor processing effects of text detection in the prior art. The method comprises the following steps of: searching for a keyword in a text segment to be processed in an inverted index, counting the occurrence number of each text segment or an identifier of the text segment in a pre-stored text segment set in a term comprising the keyword, and sequentially selecting a plurality of text segments from the pre-stored text segment set from large occurrence numbers to small occurrence numbers, wherein the inverted index is established for the pre-stored text segment set, and comprises a plurality of terms, each term comprises a keyword, and stores the corresponding text segment or the identifier of the text segment comprising the keyword; processing similarity between the text segment to be processed and each text segment in the selected text segments to obtain a plurality of similarity values; and judging whether a minimum value in the similarity values is within a set range or not, and if the minimum value in the similarity values is within the set range, outputting information comprising preset contents.

The invention discloses a portrait audio video synchronous calibration device and a portrait audio video synchronous calibration method. According to the portrait audio video synchronous calibration device and the portrait audio video synchronous calibration method, existing proven face recognition technology, dynamic lip recognition technology, voice extraction technology and the like are used, and informatization means and hardware implementation are designed, and therefore, and a portrait audio video synchronous calibration function is realized; low-time complexity left shift, right shift and XOR computation are adopted, so that computational performance can be improved; and timestamp information is not required to be added to audio and video files, and the volume of information storage can be reduced. The portrait audio video synchronous calibration device and method of the invention are applied to the synchronous detection of portrait audio videos and calibration of asynchronous audio videos.

The invention discloses a large-scale power system ill-condition load flow analysis system and belongs to the technical field of power system simulation and analysis. The system comprises a BPA data interface module, a sparse matrix calculation module for related sparse matrix processing in the load flow calculation, a system modeling module, a load flow calculation module and a load flow result analysis module. A conventional NR algorithm and a PQ decomposition algorithm are adopted to obtain a power system load flow calculation result rapidly and efficiently according to a power grid data file (*.dat) of the PSD-BPA form. The problem that part of load flow is not convergent is solved by using the PQ decomposition algorithm as an NR algorithm initial value after repeated iteration; the LM method effectively realizes ill-condition load flow solutions and analysis under the extreme operating situation.

The invention relates to detection of epileptiform activity. In order to accomplish a mechanism with improved specificity to epileptiform activity and with the capability to detect specific type of epileptic patterns, brain wave signal data obtained from a subject is decomposed into at least one predetermined subband, each subband being indicative of a specific type of epileptiform activity. The subband-specific output data obtained represents a time series of a quantitative characteristic of the brain wave signal data. At least one measure is determined for any one or more of the at least one predetermined subband, the at least one measure belonging to a measure set comprising a first measure indicative of the entropy of the subband-specific output data and a second measure indicative of a normalized form of k:th order central moment of the subband-specific output data, where k is an integer higher than three. The presence of a specific type of epileptiform activity may be detected based on the at least one measure of the respective subband.

The invention belongs to the technical field of carbon emission prediction, and in particular relates to a carbon dioxide emission prediction method comprising the steps of collecting data including the historical CO2 emission, population, GDP per capita, urbanization rate, secondary industry added value proportion, energy consumption structure, energy strength, overall coal consumption, carbon emission strength and total export-import volume; performing non-dimensionalization on the data, computing a gray association degree between each piece of data and the CO2 emission, and screening CO2 emission influence factor indexes input by the model according to the gray association degrees to achieve feature dimension reduction; using a gray prediction model GM(1,1) to predict the screened CO2 emission influence factor index; and using predicted values of the CO2 emission influence factors to serve as model inputs, and then using an improved shuffled frog leaping algorithm to optimize a least square support vector machine model for predicting the CO2 emission. The method provide by the invention has efficient computing performance and excellent global searching ability.

The invention provides a device for decomposing the characteristics of a real symmetric matrix based on circular Jacobian, which comprises a storage module, an angle resolving module, K rotating modules and a control module, wherein the storage module is used for storing and updating the elements of an N-stage real symmetric matrix A and an orthogonal matrix Q; moreover, the storage module reads the ap, the p, the aq, the q, the ap and the q of the A according to a preset circular traversal sequence, transmits the ap, the p, the aq, the q, the ap and the q of the A to the angle resolving module, reads the left multiplication element and the right multiplication element of the A and the right multiplication element of the Q and transmits the left multiplication element and the right multiplication element of the A and the right multiplication element of the Q to the rotating modules; the angle resolving module is used for carrying out angle resolving calculation on a plurality comprising the ap, the p, the aq, the q, the ap and the q and transmitting a rotation angle to the rotating modules; the rotating modules are used for carrying out rotating calculation on the left multiplication element and the right multiplication element of the A or the right multiplication element of the Q according to the rotation angle, outputting data obtained by rotation to the storage module and updating the data; the K is the degree of parallelism; and the control module is used for controlling the data reading and updating of the storage module, the angle resolving calculation of the angle resolving module and the rotating calculation of the rotating modules. By the invention, higher operational performance is realized under a reasonable hardware resource, and the contradiction between the operational performance and the resource consumption is solved.

The invention discloses a cloud computing-based pressure casting simulation control method and belongs to the technical field of pressure casting control. A pressure casting simulation control method in the prior art does not disclose a technical scheme about how to simulate a simulation model, and the prior art is narrow in application range, cannot be suitable for various pressure casting situations and is poor in simulation effect and high in usage cost. According to a cloud computing-based pressure casting simulation technology, a cloud computing mode is introduced in a pressure casting simulation technology to realize intelligent configuration of resources. Through a pressure casting cloud service platform, the learning and usage threshold of the pressure casting simulation control technology is greatly lowered. A pressure casting cloud solver is optimized according to the pressure casting simulation control method and then is packaged in a cloud service platform. The technical scheme of pressure casting simulation is detailed, feasible, wide in application range, accurate in computation, good in simulation effect, quick and convenient in use and low in usage cost.

The invention belongs to the technical field of communication, and relates to a mobile phone based method for positioning a pseudo base station, aiming to realize the positioning and tracking of the pseudo base station only by an intelligent terminal instead of an expensive professional positioning device. The algorithm measures the power information of the pseudo base station via an intelligent terminal; a smart phone uploads the measured data to the cloud and preprocesses the data at the cloud; joining the point with high signal intensity with the point with low signal intensity, and taking a median line; based on the distribution characteristic of the signal intensity, a map is segmented by the median line; and the pseudo base station is finally positioned according to the segmentation of multiple points. The positioning and tracking of the pseudo base station are realized by only a smart phone, the searching cost of the pseudo base station is greatly reduced, and meanwhile, the searching of the pseudo base station is convenient and rapid.

The invention provides an intelligent communication system based on power lines. The intelligent communication system based on the power lines comprises power line sending and receiving modules, comprehensive control terminals and multifunctional branch terminals, wherein the power line sending and receiving modules are designed to resist high voltage, universal mains supply sockets in China are adopted by the power line sending and receiving modules, the power line sending and receiving modules are applicable to various environments with the power lines, are provided with storage batteries, and can store electricity on the storage batteries when the power line sending and receiving modules are energized through the mains supply so that electricity can be used when an outage happens. The power line sending and receiving modules, the comprehensive control terminals and the multifunctional branch terminals are in modular design, and different functions can be achieved through different modes of combination. A control portion is formed by the comprehensive control terminals and the power line sending and receiving modules, and a branch control portion is formed by the multifunctional branch terminals and the power line sending and receiving modules. According to the structure, the function of simple control over electric appliances such as electric lamps on campuses, in hospitals and the like, the function of calling and the function of giving out an alarm can be achieved. An intelligent communication device circuit based on the power lines is designed according to the practical requirements and development of power line communication, the intelligent communication device circuit based on the power lines is quite good in practicability and expansibility, is an independent communication system, and is low in energy consumption, an external 12V storage battery can be adopted, and communication can be kept under the situation of outages of the power lines.

The invention provides an intelligent electric meter and an intelligent electric meter system. The intelligent electric meter comprises an Arduino chip, an electric quantity metering chip and a communication chip which are arranged in an intelligent electric meter main body and are mutually connected through a mainboard, and the electric quantity metering chip is used for metering the electricityconsumption of a user. The communication chip is used for receiving power utilization information sent by all intelligent sockets communicating with the intelligent electric meter. The Arduino chip isused for completing calculation of power utilization safety, a power utilization mode, illegal power utilization or power utilization deviation according to the power utilization information and thepower utilization quantity. The communication chip is used for sending the calculation result to the intelligent terminal and remotely keeping communication with the intelligent terminal. By arrangingthe Arduino chip in the intelligent electric meter, calculation of power utilization safety, a power utilization mode, illegal power utilization or power utilization deviation can be performed by means of the Arduino chip, relatively efficient calculation capability is achieved, and the problem that a current intelligent electric meter cannot process more real-time data due to relatively low calculation performance is solved.

A set of processors, co-processors and processor cores having a Boolean logic unit, wherein the Boolean logic unit is operable, respectively, for performing the short-circuit evaluation of Conjunctive Normal Form Boolean expressions / operations, Disjunctive Normal Form Boolean expression / operations, or both. Each processor or processor core also includes a plurality of input / output interfaces, operable for receiving a plurality of compiled Boolean expressions / operations and transmitting a plurality of compiled results, and a plurality of registers. An associated processing method including starting an operation related to a Conjunctive Normal Form Boolean expression comprising a conjunct or related to a Disjunctive Normal Form Boolean expression comprising a disjunct, evaluating the conjunct or disjunct, and selectively short-circuiting a portion of the Boolean expression.

The invention discloses a method for automatically extracting plane patches from 3D laser point cloud data of foundation, which comprises the following steps: desampling 3D laser point cloud data of foundation; 2) calculating and quantifying point attribute; 3) clustering and selecting the seed point in the parameter domain; 4) clustering spatial domain planar patch; 5) performing least square plane parameter estimation and patch boundary calculation. That invention does not need manual intervention, has excellent anti-noise property, robustness, computational performance, and is very easy toparallelize, can be used for quickly extracting planar patch from a large range of mass point cloud data, meet the real-time processing requirements and processing accuracy requirements of building indoor and outdoor three-dimensional modeling.

The present invention discloses an electronic data processing device with dual-CPU, comprising: the first CPU and the second CPU, wherein the first CPU is connected to the second CPU. The first CPU being low power type executes the first process which needs long time to execute. The second CPU with stronger computing ability in calculating than the first CPU executes the second process which requires quick arithmetic. The first CPU may output an interrupt to the second CPU in order to command the second CPU to execute the second process. The second CPU also can output an interrupt to the first CPU, thus to command the first CPU to transfer the data worked by executing the first process.

The invention relates to a processing method of an external electromagnetic source item of a three-dimensional time domain hybrid discontinuous Galerkin method, belongs to the technical field of three-dimensional time domain electromagnetism numerical solution, and relates to a three-dimensional time domain electromagnetism time domain hybrid discontinuous Galerkin numerical method aiming at the problem of the external electromagnetic source item. The method is based on an active Maxwell equation, and a construction form of time-domain hybrid discontinuous Galerkin is derived. In the source item, corresponding processing technologies are respectively provided for the impressed current density and the impressed magnetic current density, so that the method can be suitable for different boundary conditions for different problems. Compared with an existing passive time domain hybridization discontinuous Galerkin method, the improved active time domain hybridization discontinuous Galerkin method is higher in applicability. Compared with a traditional time domain discontinuous Galerkin method, the method has less global unknown quantity and remarkable calculation performance.

The invention relates to a personal identification information recognition system and method in a high-performance IP network, and belongs to the technical field of network information security and network personal privacy protection. The personal identification information recognition system comprises a feature extraction module, a data filtering and shaping module, an algorithm execution moduleand an optimization module; no manual intervention is needed; the personal identification information recognition method comprises the following steps: step 1, inputting the traffic collected by the IP network to the personal identification information recognition system in a PCAP format; step 2, extracting network feature fields to obtain a five-dimensional data group; step 3: filtering and shaping the five-dimensional data group; step 4: establishing a user behavior tree model, calculating functions VF and UVF, performing twice detection, and outputting a PII-DK and PII-VALUE lists; step 5:performing check and diffusion by an optimization module; and step 6: outputting the PII-DK list. The system and method proposed by the invention have good compatibility and expandability.

The invention provides a neural network convolution method and device. The method comprises the following steps of judging whether parameters of a to-be-input picture meet a matrix shape calculated byTensorcore or not; in response to the fact that the parameters of the picture meet the matrix shape of Tensorcore calculation, defining an operator of two-dimensional convolution; designing scheduling meeting Tensorcore calculation according to a rule defined by an operator; initializing the accumulator, and loading the input picture into a shared memory; and executing scheduling to obtain a convolution calculation result. According to the scheme of the invention, the problem that NHWC data distribution is not supported in the TVM can be solved; the calculation performance of process convolution realized by using Tensorcore is improved by more than two times than that of the current Direct calculation method of the TVM, and excellent calculation performance can be realized in the trainingand reasoning process of an AI image recognition network model.

The invention provides a blade storage device and relates to the field of data storage. The blade storage device comprises an FPGA processor and a storage array pool provided for a bus of a high-speed serial bus card; the FPGA processor comprises a high-speed serial interface processing unit and an NAND controller; the high-speed serial interface processing unit is connected with the bus of the high-speed serial bus card; the NAND controller is connected with the storage array pool. By the use of the blade storage device, the problems of low reliability, low access speed and low requirements on work environment are solved, and a blade storage is high in reliability, high in access speed and low in requirements on the work environment, and multiple blade storage devices can be quickly and conveniently made into a high-capacity blade storage system.