19 results about "Weight value" patented technology

The invention relates to an online fault detection method for a reduced set-based downsampling unbalance SVM (Support Vector Machine) transformer. At present, the research of improving the performance of an unbalance data downsampling SVM algorithm comprises upsampling and downsampling. The SVM model calculating cost of the upsampling algorithm is increased. The downsampling algorithm is selected improperly sometimes, and thus the poor classifying effect is caused. The online fault detection method comprises the following steps: (1), acquiring a vibration signal of a transformer; (2), obtaining a noise reduction vibration signal; (3), obtaining multiple groups of fault detection feature data; (4), clustering by using a K-mean algorithm; (5), figuring out a weight value of each sample; (6), establishing a majority sample reduction vector solution optimization model; (7), obtaining an SVM fault diagnosis model; and (8), inputting a sample to be tested to an unbalance SVM detector trained in the step 7, analyzing a result output from the detector to obtain a working state of the transformer, and realizing online fault detection of the transformer. The online fault detection method is used for detecting the fault of the transformer online.

The invention discloses a face recognition method and system. According to the method, the method comprises the following steps of carrying out face detection on a video stream image; after alignment,extracting a face feature vector; calculating the similarity between each frame of image and the face feature vector of the registered face feature library; establishing a tracker for each face; determining a tracker to which the human face belongs through the cross-parallel ratio of the human face position coordinate area and the weighted value of the human face similarity; and when the trackercollects face images with a preset frame number, counting a recognition result of the face tracked by the tracker, and judging a final result of face recognition according to the recognition result, the voting proportion of each group and a similarity average value. The technical scheme has the advantages of high recognition efficiency, accurate face recognition and high feature expression richness.

A method for the preparation of samples by means of a dosage-dispensing device, a weighing system with a load receiver, a processor unit, a memory unit, and an exchangeable dosage-dispensing unit. A target container is set on the load receiver. The dosage-dispensing unit and/or the target container includes electrically insulating material. The dosage-dispensing unit has the capability to change positions relative to the load receiver. While the dosage-dispensing unit is in a first position and a target container is on the load receiver, a starting weight value is determined, the dosage-dispensing unit is brought into a second position, a dosage-dispensing cycle delivers a dosage material in a predefined amount from the dosage-dispensing unit into the target container by means of the processor unit, the dosage-dispensing unit is brought into the first position, and while the dosage-dispensing unit is at rest in the first position, an ending weight value is determined.

The invention provides an associated information recommendation method and device. The method comprises the steps of obtaining m first weight values corresponding to m second-level information nodes associated with a first-level information node, wherein the m first weight values are obtained through click jump behaviors of a user, the first-level information node is an information node which is used by the user at current, and the m is a positive integer; determining associated information to be recommended to the user according to the m first weight values and preset n recommended information nodes, wherein the n is a positive integer. The associated information recommendation method and device have the advantages that since the click jump behaviors of the user are referred through the m first weight values, the problem of content focusing for recommending node information to the user through the contents of information nodes, uploaders or writers of the information nodes, types of the information nodes and the like is avoided; since the m first weight values are obtained by referring to the click jump behaviors of the user, the effect of recommending associated information to users through a big data concept is realized.

The present invention discloses a portable terminal that measures a reference tilt and a method of measuring the reference tilt using the same. The portable terminal includes a movement sensor for measuring a tilt of the portable terminal using a gravitational acceleration value, and a controller for dividing a preset time into separate time sections and measuring an acceleration value in each separate time section by using the movement sensor, calculating a standard deviation of the accelerations in each separate time section and determining one time section in which movement is minimal, and calculating a tilt of the portable terminal from an average of the accelerations in the determined one time section or a sum of weight values therein and setting the calculated tilt as the reference tilt of the portable terminal.

The invention provides an image scene multi-object mark method based on a prior condition constraint. The method comprises the steps that the region of interest of a semantic object group is determined; the multi-dimensional feature of a test image is calculated and used as a priori appearance constraint, and the pixel-level multi-dimensional feature is converted to a super-pixel-level multi-dimensional feature; the graph model structure of the region of interest of the test image is constructed; a super-pixel in the region of interest is used as a graph structure node; the adjacent relation of the super-pixel is used as the edge of the graph structure; the corresponding feature of the priori appearance constraint is converted to an edge weight value; the initial geodesic distance is calculated and used as a node weight value; geodesic propagation is carried out; in each step of propagation, the object mark of the current seed point is determined, and the geodesic distance of adjacent points around is updated to prepare for propagation of next step; and until the end of the propagation process, the object mark of each super-pixel is acquired. According to the technical scheme of the invention, the rich feature of the object is used as the priori constraint to improve the accuracy of the object mark.

The invention provides a filtering method for suspended matters in water. The method comprises the following steps: clamping a microfiltration membrane with a pair of tweezers, putting the microfiltration membrane into a constant-weight weighing bottle, transferring the microfiltration membrane to a baking oven for drying, then cooling the microfiltration membrane to room temperature, weighing the weight of the microfiltration membrane and carrying out repeated drying, cooling and weighing until the difference between weight values obtained in two weighing is less than 0.2 mg; correctly putting the obtained microfiltration membrane with a constant weight on the filter membrane pallet of a filter membrane filter, covering a matching funnel, fixing the funnel with a clip, wetting the microfiltration membrane with distilled water and carrying out continuous suction filtration; and weighing a sample for suction filtration, continuously washing the sample with distilled water three times, continuing suction filtration to remove a trace amount of moisture, taking the microfiltration membrane out, putting the microfiltration membrane in the original constant-weight weighing bottle, transferring the microfiltration membrane to the baking oven for drying, then transferring the microfiltration membrane into a drier, then cooling the microfiltration membrane to room temperature, weighing the weight of the microfiltration membrane and carrying out repeated drying, cooling and weighing until the difference between weight values obtained in two weighing is less than 0.4 mg. According to the invention, primary drying time is prolonged, drying frequency is reduced, time for drying is shortened, and work efficiency is improved.

The invention provides a directional modulation signal design method based on spatial Fourier transformation. The method comprises the following steps of establishing desired phase wave beam spatial coverage functions in ideal conditions relative to different communication models; and acquiring a phased array element weighted value according to the spatial Fourier transformation pair relation among phase wave beams and the array element weighted value of a phased array in a directional modulation signal transmitter to achieve a directional modulation signal design in different communication models. By using the directional modulation signals designed by the method, valid users can normally demodulate received directional modulation signals through common digital signal receivers. As to wiretappers in undesirable directions, their wiretapping receivers cannot demodulate useful communication information from received signals due to the relative phase relationship distortion of constellation points in a received signal constellation diagram. The invention provides a safe transmission method for transmitting communication information in wireless channels.

The invention relates to a method for performing weight analysis on main control factors of logging parameters in a water-oil-gas reservoir. The method is characterized by comprising the following steps: 1, collecting and sorting various characteristic parameters measured by each single technology in geochemical logging, collecting oil testing conclusions corresponding to each group of characteristic parameters, and then cleaning and verifying integrated data; 2, performing standardized processing; 3, writing a ReliefF characteristic selecting program by using Matlab software, and establishinga ReliefF algorithm model for geochemical logging measurement analysis; 4, inputting standardized data into the ReliefF algorithm model, and calculating weight values of the various characteristic parameters measured by the single technology by using the RelefF algorithm model; 5, setting thresholds, screening characteristic subsets, and extracting geochemical logging sensitive parameters. The method and a gray correlation analysis method have different emphases; by the method, the defects of the gray correlation analysis method in the weight analysis are overcome; the method belongs to supervised learning.

The invention discloses an aircraft damage rush repair effectiveness evaluation method: A, determining an aircraft damage rush repair effectiveness evaluation index; B, constructing an efficiency evaluation index system; C, establishing an extension evaluation matrix of the index; D, calculating the comprehensive extension judgment matrix of the index; E, calculating that weight value of the extension relative entropy of the index; F, obtaining that comprehensive weight value of the index; G, establishing that effectiveness evaluation model. This method synthesizes the information contained inthe index data and the fuzziness of expert judgment, which provides an objective way to calculate the index weight, Based on the extension theory and entropy weight method, the method of determiningthe index weight is given, which makes the determination of index weight more reasonable, objective and scientific. According to the evaluation index, the evaluation model of aircraft damage rush repair effectiveness is established, and the aircraft damage rush repair effectiveness is evaluated reasonably, scientifically and effectively, which provides an important basis for its auxiliary decision-making.

The invention provides a design method of an impeller, belonging to the computer field. The design method comprises the following steps of: establishing an initial model of the impeller according to the initial structure of the impeller; analyzing the initial model to obtain a stress value, a deformation value and a weight value of the impeller; a lightweight impel structure being obtained by optimizing the impeller by using an optimization algorithm with the objective of the lightest weight value and the constraint condition that the stress value and the deformation value satisfy the design requirements of the impeller. The design method not only realizes the lightweight of the impeller, but also provides the constraint conditions of stress and deformation, and provides an effective and reliable design method for the lightweight and optimization of the impeller.

The invention discloses an automatic weighing method and system for a drone. The automatic weighing system comprises a pressure signal acquiring module and a pressure signal processing module, whereinthe pressure signal acquiring module comprises at least one pressure sensor; the pressure sensor is arranged on a tripod of the drone; the tripod is configured to carry the drone through the pressuresensor; the pressure sensor is configured to generate a pressure signal after being subjected to pressure and transmit the pressure signal to the pressure signal processing module; and the pressure signal processing module is configured to acquire a total weight value of the drone according to the pressure signal. The drone can be automatically weighed, manual confirmation of the weights of the load cargo by the drones with different models is not needed, the efficiency of currently confirming whether the cargo loaded by the drone is overweight is improved, the labor cost is reduced, and thetime is saved.

In a method and a media system of/for generation of at least one output signal (H_PL,H_PR) from at least one input signal belonging to a second set of sound signals (M) having a related second set of Head Related Transfer Functions, in which the media system can be a TV, a CD player, a DVD player, a Radio, a display, an amplifier, a headphone or a VCR, the method includes the steps of determining, for each signal in the second set of sound signals, a weighted relation (14) including at least one signal belonging to a third set of intermediate sound signals (CHI1, CHI2) and at least one weight value (Weights); determining a first set of Head Related Transfer Functions (HRTFs) based on the second set of sound signals, the second set of Head Related Transfer Functions and the weighted relation; and transferring at least one signal belonging to the third set of intermediate sound signals by means of at least one HRTF belonging to said first set of Head Related Transfer Functions in order to generate at least one output signal belonging to said first set of sound signals. Hereby, in the end, fewer HRTFs are determined for a subsequent transfer of input signal(s) to output signal(s). Accordingly, few convolutions are required.

The present application provides a method and system for quickly retrieving similar strings, wherein the method comprises: reading an existing preset number of text entries, splitting the text entriesinto a plurality of phrases for each text entry, and assigning a corresponding weight value to each phrase; performing hashing operation on the split phrase based on the assigned weight value to obtain a first hash string corresponding to the text entry; performing collapse processing on the first hash string to obtain a second hash string whose length meets a specified condition; a prefix tree is created for the second hash string, and a string similar to a target string is retrieved from the existing preset number of text entries based on the prefix tree. The technical proposal provided bythe application can greatly improve the speed of string retrieval.

The invention relates to a searching method. The searching method comprises the steps of respectively setting one weight value for each application station and further calculating the relative percentage of downloads of specific applications in all of application of all of application stations; calculating relative downloads of specific applications according to the relative percentage of the downloads. The invention further relates to a searching system. In the searching method and the searching system, the weight values are adopted to weigh the importance degree of the application stations in the download calculating process, accordingly the downloads of one application is re-estimated in a fair mode, the degree of user's demand for one application is objectively reflected well, the searching accuracy of a search engine is improved, and accordingly the download conversion rate can be correspondingly improved.

Provided herein is a method and apparatus for heterogeneous carrier aggregation, the method including calculating interferences regarding a plurality of interference paths between an adjacent cell and serving cell; determining weighted values for the plurality of interference paths in the serving cell; calculating an accumulated value (accumulated margin or accumulated interference) regarding the plurality of interference paths based on the weighted values in the serving cell; determining a component carrier to be used in the carrier aggregation based on the accumulated value calculated in the serving cell; and performing the carrier aggregation based on the component carrier determined.

The invention provides a method and system for obtaining the optimal weight of a network chip WRR. The method comprises that an initial weight value of each WRR queue is obtained; the initial weight values of the WRR queues are traversed to obtain a corresponding maximal common divisor; and the optimal weight corresponding to each WRR queue is obtained according to the initial weight values of the WRR queues and the maximal common divisor. According to the invention, a minimal non-zero original weight is used to search for the maximal value corresponding to the maximal common divisor in a traversal manner, the traversal range when the maximal common divisor is searched is narrowed greatly, the efficiency of calculating the maximal common divisor of the original weight values is improved, the maximal common divisor of the original weight values of all the queues is searched in the traversal manner, the original weight values are divided by the maximal common divisor, and the optimal weights of the WRR queues are generated rapidly and accurately; and it is ensured that the scheduling proportion among the queues satisfies system expectation, fluctuation during WRR scheduling is reduced, the stability and reliability of the network chip are improved, and the working efficiency of development staff is improved.

The invention discloses a distribution method for leaked water in a water supply pipe network model, and belongs to the field of building water distribution in the water supply pipe network model. Themethod comprises the steps: (I), giving consideration to three factors (pipe material, pipe diameter and pipe age) affecting the leakage of pipe segments, and taking the three factors as independentvariables in a leaked water distribution model; (II), solving the leakage weight values Cijk of all pipe segments through building a three-dimensional discrete distribution model; (III), distributingthe leaked water to all pipe segments. The method gives consideration to the factors affecting the leakage of the pipe segments, can achieve the scientific and accurate distribution of the leaked water of a pipe network to all nodes of the water supply pipe network model, remarkably improves the simulation precision of the pipe network model, brings convenience to a pipe network worker to timely obtain the operation conditions of the pipe network, and is important for the improvement of the emergency processing, the real-time dispatching and the operation management for the workers of all water companies.