22results about How to "Improve storage speed" patented technology

The invention belongs to the technical field of data storage, and provides a partition-based video slice storage method and device, and the method comprises the steps: carrying out the analysis of a to-be-processed video through employing a YOLO algorithm based on the timestamp of a real-time video and the size of a fixed video, so as to construct video slices, and determining the weight corresponding to each video slice; constructing a timeline-based video partition and video slice mapping model based on the timestamps and the threshold values of the video slices; obtaining a to-be-processed video message by using the message-oriented middleware, and analyzing video slice data from the video message; and comparing the analyzed video slice data with the timestamps of the video partitions according to the constructed mapping model so as to determine a storage position. According to the method, the newly added video partitions and video fragments can be quickly determined, and the video time point can be quickly positioned.

The present invention provides a method and device for associating network assets and vulnerabilities based on fingerprints, the method comprising: detecting network assets and obtaining return messages of the network assets; generating network asset CPE fingerprints; associating the extracted CPE information with the CVE number; storing the obtained correspondence between CPE and CVE in a CPE-CVE association database; storing the data in the CPE-CVE association database in a CPE-CVE cache database; querying the cache database based on a time value, if the network asset CPE fingerprint matches the fingerprint in the CPE-CVE cache database, adding the corresponding CVE number to the network asset identified by the network asset CPE fingerprint and storing it in the network asset database. According to the solution of the present invention, the real-time and accuracy of the association are improved; rapid association query of asset vulnerabilities can be realized, and the query speed is improved.

The invention discloses a method and system for quickly writing in a Hive partition table, and relates to the technical field of big data processing, and the method comprises the following steps: obtaining to-be-written data, and carrying out data processing and automatic merging on the obtained data according to a preset rule; the method comprises the following steps: acquiring and analyzing schema information of a target table, extracting and caching a partition field list, constructing a partition creation statement, and pre-judging and creating a partition before data writing so as to realize automatic management of a Hive partition table; before the data is written into the target table, mapping fields of the to-be-written data with fields of the target table through an intelligent data mapping technology; according to the method, data to be written are written into a target table, in the process, a multi-thread parallel writing strategy is adopted, and each thread is responsible for a set number of data writing tasks. According to the method, on the premise of ensuring data consistency and integrity, the data storage speed can be remarkably increased, high-speed execution of data synchronization is realized, and the problem of low data writing efficiency of the Hive partition table in a big data scene is solved.

The invention discloses a city update obligee relationship network analysis system and method, and the system comprises a data collection module which is used for collecting various types of data related to city update, a relationship construction module which is used for constructing a city update obligee relationship network according to the collected data, and a data analysis module which is used for analyzing the city update obligee relationship network. The analysis module is used for carrying out deep analysis on the constructed obligee relation network, outputting an analysis result to a user by adopting a graph analysis algorithm and a data mining technology and a result output module, and the dynamic updating module is used for continuously monitoring the change condition of city updating related data and evaluating potential risks in a city updating project according to the analysis result of the obligee relation network. According to the method, data safety and reliability are ensured through encryption transmission and verification, a relation network is constructed through multiple factors, city updating is assisted through accurate analysis, practicability is improved through dynamic updating and risk assessment, and powerful technical support is provided for city updating projects.

The invention relates to the field of data processing, in particular to a tunnel structure monitoring and detection data standardized database construction method. The method comprises the following steps: acquiring multi-source monitoring and detection data from a tunnel monitoring system; performing cold and hot separation processing on the multi-source monitoring and detection data to obtain detection data; the cold and hot separation treatment is related to the importance degree of a tunnel structure corresponding to each deployment position in the tunnel monitoring system; selecting a plurality of groups of detection data, and distributing index hierarchical identifiers corresponding to the plurality of groups of detection data; the index hierarchical identifier is obtained by mixing at least two of a spatio-temporal index, an inverted index and a multi-level index; and according to the index hierarchical identifier, performing hierarchical storage on the multiple groups of detection data to obtain a target database corresponding to the tunnel monitoring system. According to the method, efficient data management of the tunnel monitoring system can be realized, the storage cost and the data access speed are optimized, and the query performance, the processing efficiency and the expandability of the system are also improved.

A speech transcription method based on deep learning belongs to the field of speech recognition, and aims to solve the problem of incomplete transcription text caused by inaccurate audio text processing in short time in existing segmented speech text recognition, and is characterized in that a second text is matched with a first text according to an audio time stamp, and the first text is matched with the second text according to the audio time stamp; corresponding texts corresponding to the audios of the audio intervals in the second text to corresponding text positions of the first text, and marking speaking objects for the intervals of the corresponding texts of the first text according to the text interval marks of the second text, the first text of the speaking object in the marked text interval is obtained by voice transcription.

The invention discloses a multi-area-array phased array radar target trajectory prediction method and system, and the system comprises a data collection module which is used for receiving target data detected by a multi-area-array phased array radar, a data preprocessing module which carries out the cleaning and preprocessing of the collected target data, removes an abnormal value and noise interference through a data cleaning algorithm, and carries out the prediction of the trajectory of the target. The trajectory prediction module is used for performing trajectory prediction on the preprocessed target data by using a pre-trained deep learning model, the result output module is used for outputting a trajectory prediction result to a user, the trajectory prediction result comprises graphic display, data report form and intelligent optimization, and the trajectory prediction result is obtained according to the feedback of the user on the prediction result and the effect in practical application. And the multi-source fusion module can integrate data of the multi-area phased array radar and other sensors or data sources. The target trajectory is accurately predicted through multi-module cooperation and a deep learning model, result output is visual and diversified, and the accuracy, practicability and user experience of target trajectory prediction are improved.

The present invention provides a method for predicting the geometric parameters and crashworthiness of a composite structure of a rail vehicle based on transfer learning and image recognition, establishes a finite element model of a composite energy absorption structure of a subway train, verifies the finite element model in combination with a real vehicle collision test, and calculates the collision results of the finite element model of the composite energy absorption structure under different design parameters through an experimental design method; based on the calculated collision results, the crashworthiness characteristic indicators and structural parameters are inversely calculated from the deformed image to ensure that the energy absorption structure has a better deformation pattern during the crashworthiness optimization process, broadens the application of image recognition in the crashworthiness design of the composite energy absorption structure of subway trains, and enables big data.

The invention discloses a file security storage method, electronic equipment and a storage medium, and the method comprises the steps: extracting to-be-stored file data information, and obtaining file basic information and a file relative path set; generating a master key and a file key, configuring file storage information for the to-be-stored file data information, and taking the file basic information, the file key and the file storage information as file attribute information; encrypting the master key and the file relative path set to obtain a first ciphertext; encrypting the file attribute information by using the master key to obtain a second ciphertext; encrypting the file data information by using the file key to obtain a third ciphertext; and separately storing the first ciphertext, the second ciphertext and the third ciphertext, and configuring the third ciphertext into a component container for storage according to the file storage information. According to the method, a large number of files can be stored, meanwhile, the files are subjected to security encryption and verification to ensure the security and reliability of data, the storage speed is increased, and file information can be conveniently and rapidly looked up.

The invention relates to a fault positioning method and device, a storage medium and a computer program product. The method comprises the steps that in the process of storing collected servo operation data of a servo driver to a storage unit, if it is detected that the servo driver generates abnormal servo operation data, the abnormal servo operation data are sent to a master station, and collection is stopped after a preset duration; receiving a fault operation time period sent by the master station according to the abnormal servo operation data; and sending fault servo operation data corresponding to the fault operation time period to a master station, so as to indicate the master station to carry out fault positioning on the servo driver according to the fault servo operation data. By adopting the method, fault positioning can be accurately carried out on the servo driver.

The present invention provides a high-speed data storage method and device, and relates to the field of data processing technology. The method formats an SSD according to a general format, and receives a multi-channel data acquisition instruction through a CPU; then, according to the multi-channel data acquisition instruction, a cache area is divided for each channel of data in a DDR memory, and at the same time, multiple storage cluster areas with continuous addresses are allocated for each channel of data in the formatted SSD according to the cache time granularity; then, the multi-channel data is cached in parallel to the cache area through an FPGA module; finally, based on a parallel synchronous ping-pong cache mechanism and NVME instructions, the data in the cache area is written in parallel to the storage cluster area through the FPGA module. The high-speed data storage method provided by the present invention realizes high-speed storage of data, and can directly realize data viewing, modification, and migration and exchange through the SSD on the device side, so as to manage the stored data more flexibly and conveniently.

The present invention discloses a graphite negative electrode material with good stability, a preparation method and application thereof, and relates to the technical field of graphite negative electrode material preparation. The present invention adopts melamine resin sponge as the basic skeleton of the electrode. After carbonization treatment, the skeleton has a stable three-dimensional network structure, and its unique spatial form can effectively buffer the volume change experienced by the modified graphite during the battery charging and discharging process. In addition, the appropriate carbon layer spacing of the melamine resin sponge works together with the three-dimensional buffer network to ensure that the modified graphite can fully contact with the electrolyte, thereby improving the overall performance of the battery.

The invention relates to the technical field of artificial intelligence and big data processing, and discloses an artificial intelligence-based big data analysis platform system, which comprises a data access layer, a data preprocessing layer, an AI analysis engine layer, a data storage layer, an application service layer and a monitoring operation and maintenance layer which are in communication connection in sequence. The data access layer is used for realizing multi-source heterogeneous data access and high-frequency data caching; the preprocessing layer completes data cleaning conversion and quality detection; the AI analysis engine layer realizes full-life-cycle management, intelligent analysis and result optimization and explanation of the model; the data storage layer adopts a mixed framework for classified storage and remote backup; the application service layer provides a visual display and open interface; the monitoring operation and maintenance layer guarantees stable operation of the system. According to the invention, data processing efficiency and analysis precision are improved, and data security and user experience are enhanced.

The present invention discloses a method and system for storing data with high concurrency, the method comprising S10: obtaining a data operation object from a data operation object queue according to a data write request; S20: writing the data to be written into the cache and log file of the data operation object; S30: judging whether the cache size has exceeded the set object cache threshold; S40: if not, putting the data operation object back into the data operation object queue; S50: if yes, persisting the cache data and concurrently executing other steps. The present invention can store data to a storage file or a storage repository with high concurrency by concurrently running the persistent cache data and object cache process (other step processes except step S50), thereby increasing the number of service requests and the storage speed.

The present application provides a method and device for quickly exporting a regional sub-dataset from HBase tile storage. The method comprises: obtaining a rectangular area of ​​tile data to be exported according to the longitude and latitude coordinates of two coordinate points; determining a scan boundary according to the rectangular area; reading tile data in each Region of HBase in parallel according to the scan boundary, and storing it in a global variable HashMap; storing the tile data in the global variable HashMap in a SQLite database in a read-write separation manner. The present invention realizes the function of user-defined export of tile data area and greatly improves the efficiency of exporting tile data.

The present invention provides a method for manufacturing a memory and a memory, which relates to the technical field of storage devices and is used to solve the technical problem of low storage speed and storage efficiency of the memory. The method for manufacturing the memory includes: providing a substrate, wherein a word line is arranged in a core area of ​​the substrate; forming a dielectric layer on the substrate, etching the dielectric layer, and forming a first filling hole exposing the word line; forming a barrier layer on the hole wall of the first filling hole, wherein the barrier layer located in the first filling hole is arranged to form a first middle hole, wherein the first middle hole exposes the word line; etching the word line exposed in the first middle hole to remove the first residue on the word line; forming a first wire in the first middle hole, wherein the first wire is electrically connected to the word line. By protecting the hole wall of the first filling hole with the barrier layer, the first residue can be fully etched, thereby reducing the contact resistance between the first wire and the word line, and improving the performance of the memory.

The application discloses a multifunctional analog instrument acquisition card based on a ZYNQ-7000 chip and a data processing method. The acquisition card takes a ZYNQ-7000 series main control chip as a core, a PS end is responsible for data storage and network transmission, and a PL end covers data acquisition, output and cache modules. Four-way single-channel ADC circuits are adopted for data acquisition, and the data is transmitted to the PS end by DMA after FIFO cache and conversion; a signal output module generates analog signals according to instructions of an upper computer; data transmission is realized by multi-channel DMA to be high-speed and stable; and an Ethernet module guarantees communication with the upper computer. The application utilizes the unique structure of the ZYNQ chip, and combines with the collaborative operation of various functional modules, has the advantages of modularization, high precision, high-speed processing and scalability, effectively solves many defects of existing data acquisition systems, can be widely applied to analog signal acquisition and processing in many fields, improves the efficiency and precision of data acquisition and processing, and has a great market application prospect.

The embodiment of the invention relates to a data storage and reading method and device of a neural network, and is used for storing to-be-convolved data of a neural network convolution layer to an RAM module, and the to-be-convolved data is used for carrying out convolution by adopting a 3 * 3 convolution kernel. The data storage method comprises the following steps: sequentially storing the Ath row of data of to-be-convolved data into each RAM of the first row of RAMs; storing the Bth row of data of the to-be-convolved data in each RAM of the second row of RAM in sequence; and storing the C-th row of data of the to-be-convolved data in each RAM of the third row of RAM in sequence. According to the technical scheme provided by the embodiment of the invention, the to-be-convolved data of the convolutional layer of the neural network is stored in different row and column RAMs according to the idea of row and column separation, and the storage data addresses are allocated based on the 3 * 3 convolution kernel rule, so that the storage and reading speeds of the to-be-convolved data are effectively improved.

A text error correction method for a speech recognition result belongs to the technical field of computers, aims to provide an efficient and accurate text error correction solution and improve the quality and reliability of the text result after speech recognition, and is characterized in that a text error correction model network structure comprises a BERT layer, a full connection layer, a desection layer and a correction layer, the full connection layer is used for error detection, and the correction layer is used for error correction. And weighting the loss of the convergence layer and the correction layer to obtain the final loss for training the model.

The application provides a data processing method, which is applied to an electronic device including a first storage device and a second storage device, the first storage device is UFS or EMMC, and the second storage device is RAM. The method comprises the following steps: receiving first data of a first type; storing the first data to the second storage device when the usage of the second storage device is less than a first value; storing the first data in the second storage device to the first storage device after storing the first data to the second storage device; and storing the first data to the first storage device when the usage of the second storage device is greater than or equal to the first value. In this way, by using the RAM as the cache of the first storage device with slow storage speed, the data storage speed of the storage system can be improved.

The embodiment of the invention discloses a construction operation information storage method and device, electronic equipment and a computer medium. According to one specific embodiment, the method comprises the steps that a construction data set is obtained, the construction data set comprises but is not limited to at least one of personnel data, mechanical equipment data, vehicle data, environment data and monitoring data, and the construction data set is a set of various data generated in the building construction process; performing data preprocessing on the construction data set to obtain a processed construction data set; inputting the processed construction data set into a pre-trained quality prediction model to obtain a quality prediction result; displaying the construction operation information set to a construction system so that a user can control and process the construction operation information set to obtain a controlled construction operation information set; and storing the controlled construction operation information set to a construction system. According to the embodiment, the integrity of the construction operation information is improved, the storage speed is increased, and the practicability of the construction system is improved.

The present invention relates to an image processing method, and in particular to an image fast calculation and processing method based on wave aberration correction, which solves the problem that the existing solution is slow and cannot simultaneously meet the low power consumption and low delay requirements of wave aberration correction and satellite platform use. The method comprises the following steps: Step 1: Parameter initialization: 1.1) Setting the initial matrix of wave aberration control voltage; 1.2) Setting the response function coefficient matrix, splitting and storing it in DDR3; 1.3) Setting the voltage gain matrix; Step 2: Reading the wavefront image; Step 3: Windowing and packetization; Step 4: Calculating the centroid offset matrix; Step 5: Calculating the product matrix; Step 6: Calculating the wave aberration control voltage matrix of this time; Step 7: Obtaining the wavefront image after adjusting the deformable mirror type; Step 8: Determining whether an end instruction is received; If not, returning to step 2; If so, completing. The present invention is suitable for image processing of Hartmann wavefront correction.