43results about How to "Reduce computing resources" patented technology

The invention discloses a new non-interactive K-nearest neighbor classification method under privacy protection, invention relates to the vector classification field of K nearest neighbor classification algorithm under privacy protection. The steps are as follows: 1. The client end encrypts the training data in the training data set composed of a plurality of training data with labels by a vectorhomomorphic encryption method to obtain a ciphertext data set and an intermediate matrix, and uploads the ciphertext data set and the intermediate matrix to the cloud; 2. That client end receives theplaintext vector group to be classified and encrypt the plaintext vector group to obtain the ciphertext vector group, and uploads the ciphertext vector group to the cloud end; 3. According to that ciphertext data set and the intermediate matrix, the cloud end calculates the similarity between each ciphertext vector in the ciphertext vector set and all ciphertext data contain in the ciphertext dataset, obtains the classification result set of the ciphertext vector set according to the nearest neighbor classification algorithm, and sends the classification result set to the client end. The invention greatly improves the efficiency and security of encryption, realizes non-interactive technology, achieves real outsourcing calculation, and reduces the calculation pressure of the client.

The invention relates to a minimum value calculator of an LDPC encoder and a construction method thereof. The minimum value calculator comprises a multilevel ordering unit and an inquiry unit. The multilevel ordering unit complements the ordering step in an ordering-inquiry type minimum value calculating method in a hierarchical manner; the hierarchical manner divides an input set in a recursive manner; a plurality of sets obtained in each division are processed by calculating operation on a minimum value, a secondary minimum value and minimum value serial number so as to form one level, thereby forming a hierarchical type hardware configuration which is a regular hierarchical configuration or an irregular hierarchical configuration; and the construction method of the minimum value calculator based on the regular hierarchical configuration forms the multilevel ordering unit in a stepwise manner by determining a hierarchical cardinal number. The minimum value calculator and the construction method thereof simplify the HPU design of the LDPC encoder, reduces the HPU hardware source consumption, and is easy to perform pipelining design on HPU of the LDPC encoder, thereby partitioning a key path of HPU, and increasing the working speed and the data throughput of the LDPC encoder.

The invention relates to a simulation method based on a three-dimensional simplified model of a lithium ion battery, which comprises the following steps of: (1) selecting a commercial lithium ion softpackage battery, and establishing three-dimensional layered electrochemistry-electrochemistry of three sizes of the battery according to relevant parameters; A; a thermal coupling model: a model (OUM) of one cell, a model (TUM) of two cells, and a half-sized model (HSM); (2) respectively comparing the thermal behavior and the electrochemical property of the lithium ion battery under the three simulation methods; A; and (3) discussing whether a simulation method of a full-size model can be replaced by a simulation method of a unit model, and giving a basis. According to the method, t, the three-dimensional model simulation method of the lithium ion battery can be simplified, a simulation method based on a unit model is used for replacing a simulation method of a full-size model, t, the calculation time is saved, t, the model accuracy can be ensured, and a certain guidance basis is provided for simulation researchers.

`The invention provides a group message encryption method and a device. An encryption key Ks is randomly generated and is distributed to group members when a group is established; a group message Msgis encrypted through utilization of the Ks, thereby generating a group encrypted message EMsg; the EMsg is signed through utilization of a private key of a sending client, thereby generating a first signature SignMes; a server verifies the SignMes through utilization of a corresponding pubic key; and the EMsg is forwarded to other clients in the group if the SignMes is verified successfully. The server does not know plaintext content of the group message, so the group message sending security is improved. The encryption key Ks employed each time when the group message is sent is relatively fixed, the encryption key is not generated each time when the group message is sent, the encrypted encryption key is not sent among the clients in the group, and the clients receiving the group message do not need to decrypt to obtain the encryption key, so computing resources and network bandwidth resources of the clients can be reduced.

The invention provides a method and a device for updating a virtual switch, a storage medium and equipment. The method comprises the following steps: respectively sending a forwarding rule of a virtual machine which exchanges data with the virtual switch to a virtual network card of the virtual machine and a physical network card corresponding to the virtual machine from the virtual switch; establishing a data transmission link between the virtual network card and the physical network card, so that the virtual network card and the physical network card directly forward data of the virtual machine through the data transmission link according to a forwarding rule; updating virtual switches.

The invention provides a method and device for acquiring an energy spectrum and a device for setting an energy window. A voltage comparator is in communication connection with a detector module and receives a flicker pulse signal sent by the detector module; a DAC (digital to analog converter) is in communication connection with the voltage comparator and transmits threshold information to the voltage comparator; the voltage comparator sends a flicker pulse signal that meets the threshold information to a counter; a control and storage unit is coupled to the counter and stores and records thecount information, and the control and storage unit is also connected to the DAC and stores and records the threshold information. The method comprises the following steps of: setting a threshold of the DAC and resetting the counter; using the counter to record the pulse number of flicker pulses whose amplitude exceeds the threshold within a T time period; setting the threshold of the DAC again, resetting the counter, and then recording the pulse number of flicker pulses whose amplitude exceeds the threshold within the T time period, and repeating the operation until the threshold is zero. Themethod and device for acquiring the energy spectrum and the device for setting the energy window can effectively reduce the cost and complexity of systems such as PET and SPET, and reduce the system's demand for computing resources and time.

Embodiments of the invention disclose an image processing method, a mobile terminal and a computer readable storage medium. The image processing method applied to the mobile terminal comprises the steps of determining a target area needing a pre-determined image operation in a first image generated based on a first image file, wherein a pixel in the target area is a first pixel, a pixel outside the target area is a second pixel, and the first image file is an image file stored in a first pixel storage format; performing the pre-determined image operation based on the first pixel and generatingthird pixels; ranking the third pixels, and generating fourth pixels; and generating a second image file according to the fourth pixels and the second pixels, wherein the second image file is an image file stored in a second pixel storage format.

The invention relates to a multi-agent event-driven formation control system. The system comprises a leader and a plurality of followers, wherein the leader is in communication with at least one follower and is used for generating a first follow instruction according to first action information; the followers are in communication with one another and are used for generating a second follow instruction according to second action information and the first follow instruction or generating the second follow instruction according to a third follow instruction and the second action information and realizing a preset formation according to the second following instruction. The multi-agent event-driven formation control system is a nonlinear multi-agent formation control system; compared with existing linear multi-agent formation control systems, the system has the advantages that the application scope is expanded, and the system is more conductive to practical application.

The invention provides an image coding method, which comprises the following steps that: obtaining a plurality of groups of human skeleton data for executing a target action, each group of human skeleton data comprising joint point data when the target action is executed; Extracting a motion characteristic matrix corresponding to the plurality of groups of human skeleton data according to the joint point data in the plurality of groups of human skeleton data; And encoding the motion characteristic matrix to obtain a motion characteristic image. The motion feature matrixes corresponding to themultiple sets of human body skeleton data are extracted, and then the extracted motion feature matrixes corresponding to the multiple sets of human body skeleton data are encoded into the motion feature image. Because the data volume of the motion feature image is smaller than the data volume of the plurality of action feature vector sequences, the consumption of storage resources and computing resources is reduced.

The invention relates to a data processing method and device and a storage medium. The method comprises the steps that enabling a main thread of an application to obtain attribute information and/or style information of a display element; judging whether an attribute object corresponding to the attribute information and/or the style information of the display element exists in a preset object poolor not; and when an attribute object corresponding to the attribute information and/or the style information of the display element exists in an object pool, obtaining the attribute object, and creating and generating a component object corresponding to the display element based on the attribute object. According to the embodiment of the invention, the common use of the attribute objects can be realized, so that memory resources and computing resources required when the component objects corresponding to the display elements are created are reduced, the efficiency of creating the component objects is improved, and the rendering efficiency of the display elements is improved.

The invention provides a battery anomaly detection method and device, an electronic device and a storage medium, and is applied to the technical field of battery detection. The method comprises the steps: obtaining target behavior data of a to-be-detected battery; performing self-adaptive transfer learning on a first detection model by using the target behavior data to obtain a second detection model, wherein the first detection model is obtained by pre-training according to historical behavior data of a compliant battery; obtaining a second deviation degree of the second detection model; andunder the condition that the second deviation degree is greater than a deviation degree threshold, determining that the to-be-detected battery is an abnormal battery, wherein the deviation degree threshold is obtained in advance according to a first deviation degree of the first detection model. According to the scheme, various potential abnormal behaviors of the to-be-detected battery can be effectively detected, and manpower and computing resources required by training of the battery detection model are greatly reduced.

The invention relates to a data processing method and device and a server. The method comprises: obtaining a data query request, wherein the query request comprises a data identifier; obtaining a current corresponding generation moment of the offline database; determining a target time period of the to-be-retrieved online data according to the current corresponding generation moment of the offline data; and retrieving and acquiring target data corresponding to the data identifier from the offline database and the online data in the target time period. Therefore, according to the method, the latest offline data updating moment is determined through the current corresponding generation moment of the offline database, so that the currently synchronized offline data and the real-time data in the corresponding time period can be inquired according to the latest offline data updating moment; and the offline data changed during updating can be conveniently synchronized to the offline database, so that the offline data can be queried in one full-amount offline data table, the data query task can be completed by only storing one full-amount offline data table in the offline database, the storage space is saved, and computing resources and time required by offline data synchronization are reduced.

The invention provides a mimicry security system and method based on sampling detection. The mimicry security system comprises an input agent, a redundant controller, at least two heterogeneous function equivalents and an output agent, the redundant controller is used for carrying out parameter comparison on a judgment result of the heterogeneous function equivalence body and informing the outputagent of a judgment result; the heterogeneous function equivalence body is used for selecting and synchronizing sampling data, carrying out feature judgment on the sampling data and sending a judgmentresult to the redundant controller, wherein the mimicry judgment parameters of the heterogeneous function equivalence body are sampling data generated based on the storage data; the method greatly reduces the storage space and computing resources required by the decision parameters, guarantees the timeliness and accuracy of the decision parameters, improves the mimicry decision efficiency in network edge computing scenes such as large-scale data computing, and improves the availability of a mimicry security system.

The invention discloses a method for reconstructing a flame three-dimensional temperature field of a double-light-field camera, and the method comprises the following steps: carrying out the preprocessing of a light field image, and enabling a flame region in the light field image to become a communication region; establishing a cylindrical medium model covering a flame communication area according to the preprocessed light field image; according to the established cylindrical medium model and the preprocessed light field image, carrying out flame three-dimensional contour extraction; and calculating the grids in the three-dimensional contour to finally obtain a three-dimensional temperature field. In the three-dimensional temperature field reconstruction process, only the grids in the flame need to be calculated, the reconstruction resolution of the three-dimensional temperature field is greatly improved, and the calculation time and calculation resources are greatly reduced. The method provided by the invention is high in automation degree, the shot light field image is directly operated, a preset template is not needed, and a technical basis can be provided for measuring the three-dimensional temperature field of the multi-peak complex flame by using a radiation imaging method.

The embodiment of the invention provides a method and apparatus for obtaining an illumination intensity. The method comprises: with a preset virtual light source, a preset number of virtual photons are transmitted to a target triangular patch in a virtual scene in a simulation manner; an illumination intensity at a photon detection point at the target triangular patch is recorded; according to a preset first random point selection rule, a plurality of random points are selected at the target triangular patch; on the basis of the illumination intensity at the photon detection point, a first type of illumination intensity of each random point is calculated; and according to the first types of illumination intensities, illumination intensities of all grid points at the target triangular patch are obtained. Therefore, the illumination intensities of all grid points can be obtained by the first types of illumination intensities of the random points of the target triangular patch, so that the computer resources needed by obtaining the illumination intensity of each triangular patch can be reduced and the speed of obtaining the illumination intensities of all grid points of each triangular patch after rasterization can be increased.

The invention discloses an optimum design method for a reactor core fusion collector. Modeling is conducted on a simplified reactor core fusion collector, and according to the symmetrical border, modeling is conducted on part of trays; a fusant is simplified to a cylindrical shape, the initial condition is set as complete fusion of a box of fuel elements, and the fused fuel elements fall from a big grid plate to a top cover of the reactor core fusion collector and the trays; a three-dimensional DEM method is adopted for calculation, at least ten sample points are uniformly taken from the calculated result in the radial direction, and non-dimensional unevenness sigma is defined to quantize the influences of different structures and sizes of reactor core fusion collectors on the distributionof the forming thickness of a debris bed. Through simplified modeling, the required time of DEM modeling calculation is greatly reduced, the calculation accuracy is ensured, the determination difficulty of calculation parameters is reduced, and definition unevenness provides basis for the design optimization of reactor core fusion collectors.

The invention relates to a corpus noise reduction method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring estimated label distribution of an initial corpus set; obtaining a confidence matrix according to the estimated label distribution, wherein the confidence matrix is used for describing label noise distribution under a category condition; obtaining a noise corpus in the initial corpus set based on the confidence matrix; and processing the noise corpus in the initial corpus set to obtain a target corpus set. In the embodiment, the confidence matrix can be established through the prediction probability of the label and the labeling label, the noise corpus in the initial corpus set is identified through the confidence matrix, after the noise corpus is processed, the proportion of the noise corpus in the target corpus and ambiguous information can be reduced, the boundary of the target corpus is clearer, and the training frequency of the vertical domain model is reduced, so that computing resources and consumed time required by training are reduced, and the training efficiency is improved.

The invention relates to the technical field of engineering design, in particular to an engineering design method for conducting automated evolution design on mixing field systems or various different field systems based on a bond graph tool and a genetic programming algorithm. The method includes the steps of assigning an embryo physical structure diagram, conducting user input, setting up an initial population of a genetic programming tree, conducting ontoanalysis, conducting evolution operation and ending judgment. Compared with the prior art, the design method is high in expansibility, wide in application range, high in efficiency, high in normalization and capable of designing a complex project meeting the requirements of a designer through an automated method.

The invention discloses a block chain fine-grained editing method supporting rapid state updating, and the method comprises the steps: for each transaction executed by an editable block chain for the first time, generating a rapid execution code of the transaction according to the execution process information of the transaction, and recording a read set and a write set of the execution process of the transaction; after an editing event that a block chain editing authority owner edits a certain historical transaction on a block chain occurs, transactions of which the states of read set elements are changed compared with those in the first execution are optimally executed in parallel, and the execution of the transactions is accelerated by using a quick execution code generated in the first execution of the transactions. And after all transactions in a certain block B are updated, state tree, transaction tree and receipt tree information in the block heads of the block B and all blocks which are influenced by the update before are updated in parallel. According to the method and the device, the time and computing resources required by updating the state of the block chain node after the editable block chain node performs the editing operation can be reduced, and the performance of the editable block chain is improved.

The invention discloses a satellite component layout temperature field prediction method based on a physical prior neural network, and the method comprises the steps: building a structural model of the satellite component layout according to the layout characteristics of a satellite component; acquiring a plurality of training data, and preprocessing the training data; according to a heat conduction steady-state equation obeyed by a satellite component layout temperature field, constructing a loss function embedded with physical prior; the weight of each prediction point of the temperature field is determined through online data mining, a loss function is updated according to the weights of the prediction points, a regularization item of the loss function is constructed through regional heat flux conservation, and a final loss function is determined; constructing a deep neural network model, and training the deep neural network model by using the preprocessed training data and the final loss function; and utilizing the trained deep neural network model to predict the temperature field of the satellite component layout. According to the method, stable and rapid training of the deep neural network model can be realized by using the training data without labels, and the prediction precision of the model is ensured.

The embodiment of the invention provides a method for obtaining object track similarity. The method comprises the following steps: obtaining a spatial-temporal characteristic matrix corresponding to each target object; performing dimension reduction operation on the spatial-temporal characteristic matrix to obtain a signature matrix corresponding to the spatial-temporal characteristic matrix; performing similarity calculation on the signature matrixes corresponding to any two spatial-temporal feature matrixes to obtain the signature matrix similarity between any two target objects; and selecting two target objects of which the signature matrix similarity is greater than a preset similarity threshold, and determining the trajectory similarity between the two selected target objects based on the spatial-temporal feature matrixes corresponding to the two selected target objects. According to the method, dimension reduction operation is carried out on the target spatial-temporal characteristic matrix corresponding to the target object, so that the dimension of data is greatly reduced, the data volume is greatly reduced, the track similarity obtaining efficiency of the target object is improved, the tracking response speed of the tracked object is high, and the real-time performance is good.

The invention discloses a power grid quasi-real-time intelligent control method and system, and the method comprises the steps: obtaining the observed quantity of a power grid state, wherein the observed quantity of the power grid state comprises the unit output in the current time period, a load active power predicted value in the next time period, a load active power change value of the next time period relative to the current time period, a line on-off state, a unit output adjustable range upper limit and a unit output adjustable range lower limit; inputting the observed quantity of the power grid state into the trained intelligent agent, taking the output result of the trained intelligent agent as a unit adjusting instruction to adjust each unit in the power grid, and completing the power grid quasi-real-time intelligent control, wherein an actuator loss function of the intelligent agent in the training process is constructed based on a scheduling experience regular term. According to the method and the system, quasi-real-time control of the power grid can be realized.

The invention is suitable for the technical field of communication, provides a data transmission control method based on an RS485 bus and a related device, and realizes reduction of waste of computing resources and cost of a host. The method is mainly applied to an RS485 bus control system, one end of the RS485 bus control system is electrically connected with a slave computer, the other end of the RS485 bus control system is electrically connected with X hosts through an RS485 bus, and X is a positive integer larger than 1. The data transmission control method comprises the steps that a target data request is received, where the target data request is a data request sent by a target host to a slave, and the target host is one or more of X hosts; the target data request is only forwarded to the slave, so that the slave processes the target data request to obtain corresponding target response data; the method also includes receiving the target response data fed back by the slave; and only forwarding the target response data to the target host.

The invention provides a security and protection equipment active defense method, device and system. The method comprises the steps that security and protection front-end equipment acquires first equipment operation state information in preset unit time; the security and protection front-end equipment determines a first statistical characteristic value of the first equipment operation state information by using a determined change point detection algorithm based on the first equipment operation state information; the security and protection front-end equipment determines whether the current equipment operation state is abnormal or not based on the first statistical characteristic value and the determined second statistical characteristic value; and when the security and protection front-end equipment is currently in an abnormal operation state, an alarm is given to the equipment management platform. According to the method, computing resources and storage resources required for realizing active defense of the security and protection front-end equipment can be reduced.

The invention relates to a data reading method and device in a host performance acceleration mode. The method is executed by a host end, and comprises the following steps: searching a host performance acceleration buffer area in a system memory to obtain a logic physical comparison record associated with a logic block address; sending a switching command to a flash memory controller for requesting the flash memory controller to start a host performance acceleration function, but not start an acquisition function of a logic physical comparison table; sending a multi-block writing command to the flash memory controller for writing data blocks into the flash memory controller, wherein the data blocks comprise logic physical contrast records; and sending a command for reading the plurality of blocks to the flash memory controller for acquiring data corresponding to the logical and physical contrast record from the flash memory controller. According to the invention, the host end can send the read command with the logic and physical comparison record to the flash memory controller through the arrangement of the host performance acceleration buffer area, so that the time and operation resources spent by the flash memory controller are reduced to carry out logic and physical comparison conversion.

The invention discloses a pre-training method and device, equipment and a storage medium, and belongs to the technical field of computers. The method comprises the following steps: obtaining an initial text sentence after character masking processing; obtaining a target text sentence based on the initial text sentence after the character masking processing and the additional characters before the sentence; a mask matrix corresponding to the target text sentence is determined, the mask matrix comprises a plurality of elements, each element is used for indicating the operation association degree of two characters corresponding to the element in the target text sentence in the feature extraction process to a to-be-trained feature extraction model, and the element corresponding to the additional character in front of the sentence is not 0; and training a to-be-trained feature extraction model based on the initial text sentence, the target text sentence and the mask matrix. By adopting the method and the device, not only can the feature vector corresponding to each character in the target text sentence be obtained, but also the feature vector corresponding to the target text sentence can be obtained, other training does not need to be carried out additionally, and data operation resources and operation time are reduced.

The embodiment of the invention provides an image retrieval method. The retrieval method comprises the following steps: acquiring a first image to be retrieved, and inputting the first image into a trained image representation network to obtain a first representation vector; carrying out binarization processing on the first representation vector to obtain a binary representation vector of the first image; on the other hand, obtaining a plurality of binary representation vectors of the plurality of candidate images; and further, respectively calculating vector distances between the plurality of binary representation vectors and the binary representation vector of the first image, and recalling an image similar to the first image from the plurality of candidate images based on the vector distances. The embodiment of the invention further provides a training method of the image representation network. The training method can ensure that the retrieval performance of the image features before and after binary processing is kept consistent.

The invention provides a small hydropower station group short-term robust optimization scheduling method considering prediction errors. The method comprises the following steps: acquiring samples of uncertain factors after the prediction errors are considered by using a large amount of historical data, wherein the samples are subjected to numerical optimization processing, and an output set of the runoff type small hydropower stations is constructed; and constructing a short-term robust optimization scheduling model, and substituting the output set into the short-term robust optimization scheduling model for solving to obtain final output of the small hydropower station. According to the optimization scheduling method, samples of prediction errors of the runoff type small hydropower stations are included in the uncertain set, the prediction errors are considered while short-term optimal scheduling of the small hydropower station group is achieved, model solution conservative property is reduced, and robustness of the uncertain set and power generation benefits of the power stations in the area are improved.

The invention relates to a data reading method and device in a host performance acceleration mode. The method is executed by a host end, and comprises the steps: obtaining a value of a dedicated data register of an expansion device in a flash memory controller from the flash memory controller; and when the value of the expansion device dedicated data register contains the information supporting the host performance acceleration function, configuring a space in the system memory as a host performance acceleration buffer, and storing a plurality of logical and physical contrast records acquired from the flash memory controller in the host performance acceleration buffer. The host end and the flash memory controller communicate with each other by using an embedded multimedia card communication protocol, and each logic physical contrast records information of data of a storage logic address actually stored in a physical address. According to the invention, the host end can send the read command with the logic and physical comparison record to the flash memory controller through the arrangement of the host efficiency acceleration buffer area, so that the time and operation resources spent by the flash memory controller are reduced to carry out logic and physical comparison conversion.

The invention discloses an optimization method for STATCOM transient voltage control instruction, which comprises the steps of obtaining filtered electrical quantity data when it is detected that an STATCOM coordination controller operates in a constant reactive power control mode; storing the filtered electrical quantity data in a first register and a second register, and updating the filtered electrical quantity data and corresponding updating moments at preset time intervals; when it is detected that the STATCOM coordination controller is switched from the constant reactive power control mode to the transient voltage control mode, stopping updating, and extracting filtered electrical quantity data in the target register according to the switching moment and the updating moment currentlystored in the first register and the second register; and according to the filtered electrical quantity data in the target register, calculating a voltage instruction value of STATCOM transient voltage control, controlling the STATCOM coordination controller to operate in a transient voltage control mode, and improving the stability of power grid operation. The invention further discloses a device and equipment for optimizing the STATCOM transient voltage control instruction and a storage medium.