40results about How to "Uniform data" patented technology

A sensor module for a sensing apparatus configured for operation downhole, within a borehole. The sensor module comprises a sensor array having a plurality of sensors to sense selected formation parameters and a control system for selective and independent operation of each sensor of the sensor array. Each sensor of the sensor array is configured or designed as a discrete sensor unit for individual and independent communication and control. Each sensor of the sensor array may have an associated electronics module that provides standardized electronic connectivity with the control system.

Disclosed is a transformer fault detecting method based on a simplified set unbalanced SVM. The method comprises (1) obtaining a characteristic vector set through a fault characteristic extracting method based on GARCH models; (2) performing determination of boundary samples on minority-class samples to obtain a minority-class boundary sample set S, wherein the minority-class samples are fault samples; randomly selecting N[x]={2, , ISI}, wherein ISI is the cardinal number of S, a[i]=1, and i=1, N[x], and setting N[z] to be 1, utilizing a simplified set solution algorithm to obtain Z[1] and repeating the operation for N[L]-N[M] times, wherein N[L] is the number of majority samples, N[M] is the number of minority samples, and accordingly, N[L]-N[M] is the number of artificial minority samples, and guaranteeing N[z]=ISI for at least one once; (3) combining the artificial minority samples obtained in the step (2) with original minority samples to serve as the training samples of an SVM classifier and lastly to obtain an SVM decision model; (4) inputting newly-obtained transformer characteristic vectors into the decision model for judgment. The transformer fault detecting method based on the simplified set unbalanced SVM is applied to transformer fault detection.

The invention discloses a data equalizing method and device for a server cluster. The specific implementation mode of the method comprises the steps that expansion on the server cluster is responded, and at least source disk storing to-be-migrated data and a target disk for storing the to-be-migrated data are determined, wherein the source disks are located in at least one original server of the server cluster before expansion, and the target disk is located in a new server in the server cluster obtained after expansion; the new server is forbidden to receive a read-write request on the target disk, and a migration operation of migrating the to-be-migrated data from the source disks to the target disk is executed; when the target disk meets the preset condition that the remaining capacity of the target disk is the same as the average disk remaining capacity of the source disks, the migration operation stops being executed, wherein the average disk remaining capacity is obtained by dividing the sum of the source disk remaining capacities by the number of the source disks. The purpose that the expanded server cluster reaches data equalizing in a short time is achieved, and the cost is reduced.

The invention discloses a data link equalization method, a mobile terminal and a computer readable storage medium. The data link equalization method is applied to the mobile terminal including two WiFi antennas and two data channels, and used for the data links generated by the upper application and transmitted by two data channels when two wireless hotspots are connected at the same time. The data link equalization method comprises the following steps: detecting the link delay of the two data channels by the mobile terminal; calculating the average link delay of the two data channels based onthe detected link delay and the preset smoothing algorithm; and confirming the data link equalization ratio of the two data channels according to the average link delay of two data channels, so thatthe data links newly built by the upper application can be allocated to two data channels for transmission based on the confirmed data link equalization ratio when the data links newly built by the upper application are detected. According to the invention, the data link equalization ratio is adjusted through the link delay of the two data channels, therefore, the data balance of the two data channels can be effectively guaranteed, and the network access efficiency is improved.

The present invention relates to the technical field of computers, and provides a data control method and device, a computer device and a storage medium. The method comprises the steps of: obtaining data to be accessed in a target interface access request, according to a token bucket algorithm, calculating target access data which are allowed to access by the target interface, recording actual flow data when the target access data passes through the target interface, comparing the actual value of each flow parameter of the actual flow data with a standard value, marking the flow parameter withthe actual value being larger than or equal to the standard value and a predicted value, and if the difference is larger than a preset target threshold value, determining safety flow data corresponding to the difference as abnormal flow data. The data control method and device, the computer device and the storage medium can be improved in the detection accuracy of the abnormal flow data.

The invention discloses a data link allocation method, a mobile terminal and a computer readable storage medium. For a mobile terminal including two paths of WiFi antennae and two data channels, whentwo wireless hotspots are connected at the same time, the two data channels transmit data links generated by an upper application, the mobile terminal detects real-time transmission rates that the twodata channels transmit the data links, an average transmission rate of the two data channels is calculated based on the detected real-time transmission rates and a preset smoothing algorithm, and then a data link allocation ratio of the two data channels is determined according to the average transmission rate of the two data channels, and thus when detecting that the upper application newly establishes data links, the data links newly established by the upper application are allocated to the two data channels for transmission based on the determined data link allocation ratio, and the data link allocation ratio is adjusted through the transmission rates of the two data channels. Thereby, the data balance of the two data channels can be effectively ensured, and the network access efficiency can be improved.

The invention discloses a method of solving data imbalance based on Epochs and belongs to the field of deep learning. In training process, each Epoch randomly resamples each class according to weightso that samples in each Epoch can be averagely represented during training; addition is made to each sample according to the resampled weights, a sample set of single Epoch size is randomly resampledfrom a sample base according to a weight ratio so that data of the resampled Epochs are relatively balanced. The method has the advantages that the data imbalance problem can be more effectively solved; in the training process, each Epoch randomly resamples each class according to the weight so that the samples in each Epoch can be averagely represented during training; the main idea lies in thataddition is made to each sample according to the resampled weights, the sample set of single Epoch size is randomly resampled from the sample base according to the weight ratio, and accordingly, dataof the resampled Epochs are relatively balanced.

A motion picture data transmission method for transmitting motion picture data from a server system via a transmission line to a client system is disclosed. The content of the transmitted motion picture data is checked and the transmission interval in each unit time is adjusted on the basis of the content data of the motion picture data. Thus, it is possible to eliminate missing of data, delay of transmission data, etc., resulting due to lack of uniformity of the transmission data quantity on a transmission line for transmitting motion picture data from a server system to a client system.

The embodiment of the invention provides a method and a device for expanding training samples. The method comprises: acquiring a to-be-expanded initial training sample group, the to-be-expanded initial training sample group comprising a first number of training samples, the training samples comprising original machine statements and original user statements in a historical dialogue group and category labels corresponding to the dialogue group, and each training sample in the initial training sample group having a first category label; obtaining a second number of training samples from the initial training sample group; for each dialogue group in the second number of training samples, inputting a first machine statement related to the original machine statement in each dialogue group into apre-trained dialogue generation model of a first class label, and generating a first user statement corresponding to the first machine statement in each dialogue group; and adding the first machine statement and the corresponding first user statement in each dialogue group into the initial training sample group as extended training samples to obtain an extended training sample group. Data equalization of training samples can be realized.

A flash memory device and a method of manufacturing the same, wherein a silicon layer having a micro grain is formed between a tunnel oxide layer and a floating gate using a hemi-spherical grain (HSG) method, thereby preventing the dopant of the floating gate from being diffused into the tunnel oxide layer. According to one embodiment, the flash memory device includes isolation structures formed in predetermined regions of a semiconductor substrate, for defining an active region and a field region, a tunnel oxide layer formed on the semiconductor substrate of the active region, and a floating gate formed in a predetermined region on the active region to overlap with a part of the isolation structure, an underlying given portion and the remaining portions of the floating gate having different grain sizes.

The invention discloses a disk bad block processing method and device, and the method is applied to a distributed storage file system, and comprises the steps: firstly determining a data block identification corresponding to a data block with a bad block when the bad block is detected to exist in a data block in a target disk in the distributed storage file system; then, obtaining associated datacorresponding to the data block identifier according to the data block identifier; and finally, recovering lost data according to the associated data, and storing the recovered data into a reserved data block area of the target disk. According to the invention, a reserved data block area is added to each disk in the distributed storage file system; when it is detected that the bad blocks occur inthe data blocks in the disks, data recovery is conducted on bad block data, after data recovery is conducted, the recovered data is stored in the data blocks in the reserved data block areas in the disks corresponding to the bad blocks, other disks do not need to be occupied any more, and therefore data balance on all the disks is guaranteed.

The invention provides a storage system capacity expansion method and device, equipment and a machine readable storage medium, and the method comprises the steps: calculating and generating migrationinformation according to a preset rule, wherein the migration information comprises codes of data units, needing to be migrated to all empty hard disks, in all data hard disks, and a mapping relationship between each data unit to be migrated and the corresponding immigrated empty hard disk; migrating each data unit needing to be migrated in each data hard disk into the corresponding empty hard disk according to the migration information; updating the mapping relationship between the data units in a mapping table and the hard disk according to a migration result; exchanging and installing a part of originally loaded empty hard disks of newly added storage equipment and a part of originally loaded data hard disks on the original storage equipment, so that the storage equipment comprises a plurality of empty hard disks and a plurality of data hard disks. Through the technical scheme of the invention, the efficiency of completing data equalization after capacity expansion of a storage polegroup is greatly improved.

The invention provides a traffic signal lamp number identification method and system, and relates to the technical field of vehicle automatic driving. The traffic signal lamp digital recognition method comprises the following steps: acquiring an image or a video of a traffic signal lamp; identifying the shape of the traffic signal lamp in the image or video according to the image or video of the traffic signal lamp; selecting a pre-established identification model according to the shape of the traffic signal lamp; performing data preprocessing on the image or the video according to the identification model; and inputting the preprocessed image or video into a recognition model to recognize countdown numbers in the traffic lights. According to the traffic signal lamp number identification method, the identification process can be simple, countdown number identification of the traffic signal lamp is accurate, safe and stable driving of an automatic driving vehicle is guaranteed, and serious consequences of traffic accidents caused by sudden braking, red light running and the like of the automatic driving vehicle are prevented.

The invention discloses a safety helmet wearing detection method, device and equipment based on deep learning, and a medium. The method comprises the following steps: obtaining a to-be-detected person image; inputting the image of the person to be detected into a trained multi-scale sensing network model; the multi-scale sensing network model comprises three parallel CNN sub-networks, structures of the three CNN sub-networks are different in convolution kernel size, and each CNN sub-network comprises a channel attention module used for extracting global features and a space attention module used for extracting local features; and outputting an image classification result of the to-be-detected person, wherein the classification result comprises a condition that the person does not wear a safety helmet and a condition that the person wears the safety helmet. The method has a high detection rate and a low false detection rate in a complex scene of a transformer substation, and the deep learning method can reduce the previous manual monitoring operation, reduce the labor cost, and ensure the real-time performance of safety helmet detection.

The invention discloses a kudu database data equalization system based on size and an implementation method, and relates to the database field, the technical problem to be solved by the invention is how to realize size-based database data equalization of kudu. According to the adopted technical scheme, the system structurally comprises a data equalization condition detection module, a to-be-migrated Tablet selection module and a Tablet migration execution module; the data equalization condition detection module is used for detecting whether data equalization operation is executed or not; ; theto-be-migrated Tablet selection module is used for selecting a to-be-migrated Tablet and a migrated node; and the Tablet migration execution module is used for executing actual data migration. The invention also discloses a kudu database data equalization method based on size.

The invention discloses a method for solving unreasonable spatial and temporal distribution of dust load equipment. The method comprises the following steps: dividing a road into n road sections with uniform length; for each road section, uploading a group of data of each road section by the communication module to the Internet platform at a fixed frequency f each time; taking the data volume uploaded in the time period T as a data source, and calculating a standard value of the data volume of each road section; calculating the ratio of the standard value to the data volume of each road section to obtain a correction coefficient, and obtaining the corrected uploading frequency fan = f * xan; and when the positioning module recognizes that an road section is entered, transmitting a correction coefficient by the internet platform to the communication module, adjusting the data uploading frequency by the data processing module to fan, and finally presenting the corrected data volume of the road section. According to the technical scheme adopted by the invention, no extra hardware cost is added, and the method is easy to implement, therefore, road data balance is achieved, and a positive guidance value can be provided for urban governance.