195results about How to "Solve computing problems" patented technology

The invention relates to a method and a system for navigating indoor paths of maps across floors. The method comprises the following steps: receiving a request for navigating a path which comprises a map starting point and a map target point; inquiring a connection matrix of a plurality of maps of floors according to a floor of the map starting point and a floor of the map target point, and determining a connection relation between maps of the floors of the map starting point and the map target point; generating a shortest path between path points of maps of connected floors between the map starting point and the map target point, and taking the shortest path as a subpath of maps of the connected floors according to a connected map connecting table if the map starting point and the map target point are not on the same-floor map; combining all the subpaths after generating all the subpaths to form a navigation path between the map starting point and the map target point. Through the method and the system, the problem of calculating the path under indoor multi-map environment is solved, thus the requirement of a current indoor map service system on calculating paths across the maps can be met.

The invention discloses a recognition method and device for application search intentions and an application search method and server. The method includes the steps that search terms in query sessions are acquired from a query session log of an application search engine; label systems of the search terms are mined according to the search terms in the query sessions and preset strategies; the application search intentions corresponding to the search terms are recognized according to the label systems of the search terms. In the scheme, by proposing the label method of the recognition method for the user intentions corresponding to the app label systems, the fine-grained user query intentions are flexibly expressed. The label systems of the user intentions are constructed based on the unsupervised machine learning technology, a traditional user intention classification method is abandoned, an automatic user intention mining process is achieved, and a high-accuracy and high-recall-rate user intention label list can be generated; the user intentions and apps are mapped into the same label system, and thus a user can rapidly and accurately acquire apps meeting intentions when searching for apps.

The invention relates to a method for designing a workpiece model for manufacturing aircraft sheet metal components. The method comprises the following steps: firstly extracting a neutral surface of the component, dividing into finite element meshes, and adopting the method of directly projecting to a plane of a web along the normal vector direction of the web for calculating the node coordinates of mesh units after flattening for non-web structural elements in the web profile; and adopting the constrained surface flattening algorithm based on unit equal deformation for calculating the node coordinates of the mesh units after the flattening for the web and curved edges to be flattened. The method faces to the key manufacturing procedure of the sheet metal components and provides a data source for the molding technology and the tooling design. The method can rapidly and precisely calculate the shape of the workpiece model for manufacturing the sheet metal components, thereby meeting the needs of digital manufacture. The method can effectively solve the problems of discharge and molding workpiece calculation in the digital manufacturing of the aircraft sheet metal components.

The invention discloses a method for extracting the width of river channels within a basin by using a digital elevation model. The method comprises the following main steps: the digital elevation model (DEM) data of the basin are extracted; the slope grids, river channel grids and confluence accumulative grids of the basin are calculated according to the digital elevation model (DEM) data of the basin; near-river channel slope grids are extracted by using the slope grids and the river channel grids; near-river channel confluence accumulative grids are extracted by using the confluence accumulative grids and the river channel grids; moving average calculation is performed on the basis of the near-river channel slope grids, so that moving averaged near-river channel slope grids can be obtained; and the width of the river channels within the basin is extracted on the basis of the near-river channel confluence accumulative grids and the moving averaged near-river channel slope grids. The method of the invention has the advantages of stable and reliable data sources, high calculation efficiency, objective and reasonable results and the like. With the method adopted, the rapid calculation and extraction of the width of the river channels within the basin can be benefitted.

A method for implementing the formula editing, calculating and managing on PDA includes invoking the formula managing module to display the main interface which is composed of such function keys as increase, delete, edit, naming, calculation, preserve and exit, and clicking one of said function keys to implement relative function, such as edit and calculation for invoking the formula editing module and calculating module.

The invention provides a method for air change rate measurement in dwellings based on carbon dioxide produced by people. The method includes the steps of placing a CO2 concentration measurer, and drawing a planar space graph; recording personal information of dwellers and opening and closing conditions of doors and windows of a measured space; enabling the CO2 concentration measurer to work continuously for more than 24 hours; partitioning the measured space into one or more regions; calculating volumes of the regions; calculating the amount of CO2 produced by the people in every region according to height, weight and activity types of the dwellers; assuming an initial value of air change rate for each region; changing the assured air change rate of each region, and fitting a CO2 concentration curve so as to obtain an air change rate result of each region finally. An existing method for measuring the air change rate by tracer gas is limited economically and practically, and a method taking CO2 as the tracer gas is limited in single-region buildings. The method for the air change rate measurement has the advantages that by effective assuming, calculation for the air change rate measurement can be applied to multi-region spaces, the calculation problem for the air change rate of the multi-region spaces is solved, and the real-time air change rate of the multi-region spaces can be measured simply, conveniently, quickly and effectively.

The invention provides a method and a device for training a multi-label classification model, which can dynamically learn image features, enable a feature extraction network to better adapt to task requirements, and are good in multi-label classification effect. The method comprises the steps of determning n samples and a label matrix Yc * n corresponding to the n samples in a training data set, wherein an element yi * j in the label matrix Yc * n represents whether an ith sample contains an object indicated by a jth label or not, and c represents the number of labels related to the samples; extracting a feature matrix Xd * n of the n samples by using a feature extraction network;utilizing a feature mapping network to obtain a prediction tag matrix of the feature matrix Xd * n, wherein theprediction tag matrix is obtained by using the feature mapping network; and obtaining a prediction tag matrix of the feature matrix Xd * n. Elements in the DA00033902000001. TIF are shown in the specification. The invention also discloses a preparation method of the optical glass. The optical glass is characterized in that the optical glass is prepared from the optical glass, the optical glass and the like; a confidence representing that the i-th sample comprises an object indicated by the j-th tag; According to the label matrix Yc * n and the prediction label matrix, obtaining a predictionlabel matrix Yc * n according to the prediction label matrix Yc * n and the prediction label matrix Yc * n, and updating the weight parameter Z and the feature mapping matrix Mc * d, and training themulti-tag classification model.

The invention discloses an electric property prediction method of a cylinder conformal array antenna under load distortion. The method comprises following steps of determining structure parameters and electromagnetic parameters of the cylinder conformal array antenna; carrying out structure load distortion analysis; calculating new positions of array elements after distortion; building an array element rectangular coordinate system and an array element sphere coordinate system, and a conversion matrix between the two coordinate systems; solving an array element directional diagram under the array element rectangular coordinate system; determining a conversion matrix between the array element rectangular coordinate system and an array rectangular coordinate system; solving an array element directional diagram under the array rectangular coordinate system; determining array element execution amplitudes and phases; calculating the space phase difference of each array element; calculating a far region field intensity distribution directional diagram; building an electromechanical coupling model; calculating the electric property parameter of the antenna; and judging whether the calculated electric property parameter satisfies a demand or not. The method can be used for quantitatively evaluating the influence of antenna structure distortion on the electric property of the cylinder conformal array antenna under a load environment, thus instructing the structure design of the cylinder conformal array antenna.

The invention provides a method and a device for training a multi-label classification model, which can dynamically learn image features, enable a feature extraction network to better adapt to task requirements, and are good in multi-label classification effect. The method comprises the steps of determining n samples and a label matrix Yc * n corresponding to the n samplesfrom a training data set,wherein an element yi * j in the label matrix Yc * n represents whether an ith sample contains an object indicated by a jth label or not, and c represents the number of labels related to the samples;extracting a feature matrix Xd * n of the n samples by using a feature extraction network; using a first mapping network to obtain a prediction label matrix of the characteristic matrix Xd * n, and using a second mapping network to obtain a low-rank label matrix of the label matrix, wherein the prediction tag matrix and the low-rank tag matrix are defined in the specification; updating the weightparameter Z, the feature mapping matrix Mc * d and the low-rank label correlation matrix S, and training the multi-label classification model.

The invention provides a network function end node prediction method based on cascading failure. The method comprises steps that 1, data pre-processing on the infrastructure network is carried out, and actual network abstraction is carried out to establish a network model; 2, an initial weak node is determined based on a key node identification method or historical data, and a load capacity model is further established; 3, network segmentation points during cascading failure are calculated; and 4, a propagation distance of a function end node is predicted according to an overload cascading failure propagation distance. The method is advantaged in that the function end node in a cascading failure process can be discovered in advance at a protection stage before cascading failure, a key node or a not-easy-to-restore node is designed and distributed at the function end node in advance, through propagation prediction of the function end node, real-time control in the cascading failure process is carried out, and thereby cascading failure control and repair work afterwards are facilitated.

The invention discloses a self-checking system and a self-checking method for a cascaded multilevel converter. The self-checking system for the cascaded multilevel converter comprises a cascaded multilevel converter and a self-checking device, wherein the cascaded multilevel converter comprises at least two converting circuits which are cascaded, the self-checking device comprises at least one current detecting circuit, a voltage acquisition module and a computation module; the current detecting circuit is used for detecting a first detection circuit and a second detection circuit, the voltage acquisition module is used for acquiring a first busbar voltage and a second busbar voltage, and the computation module is used for acquiring insulation resistance of the cascaded multilevel converter according to the first detection current, the second detection current, the first busbar voltage and the second busbar voltage. The self-checking system provided by the invention can detect the insulation resistance of the cascaded multilevel converter, so that too heavy leakage current due to too low integral insulation resistance is avoided, and potential safety hazard can not be caused.

The invention relates to a satellite fast high-precision extrapolation method of a great-eccentricity track. A task track of a martian detector is an elliptic track with great eccentricity (e is greater than 0.6), a numerical integration method is adopted, and in order to realize the high-precision track calculation, the consideration of higher-order martian shape perturbation and the influence of other perturbation factors is needed, so the calculation quantity is greater, and the numerical integration method can be interrupted when a satellite computer resets or is switched; an analytical method is only applicable to the elliptic track with smaller eccentricity (e is smaller than 0.6); and the detector position provided by a precise track is utilized for fitting the Chebyshev multinomial coefficient, so greater fitting errors can be brought. The satellite fast high-precision extrapolation method has the advantages that the position difference of the precise track and the two-body track is introduced, the position difference is utilized for fitting the Chebyshev multinomial coefficient, the calculation problem of the satellite track of the elliptic track with great eccentricity can be perfectly solved, the calculation quantity is small, and in addition, the precision is high.

The invention discloses a malicious-software characteristic clustering analysis method and system based on behavior segment sharing. The malicious-software characteristic clustering analysis method comprises the following steps of: distributing acquisition and analysis nodes, establishing a distribution type hash table module, segmenting behaviors of an acquired malicious software sample into behavior segments, and storing the behavior segments to a hash table; and carrying out statistics on the global characteristics of the segments, returning the global characteristics to the acquisition and analysis nodes which construct a characteristic vector of the malicious software sample, carrying out clustering, and extracting the characteristics and the attributes of the clustering as a comprehensive analysis result and inputting the result. The system comprises a plurality of acquisition and analysis nodes which comprise a behavior-segment segmenting module, the distribution type hash table module, a behavior-segment synergy sharing module, a malicious-software sample representation module, a malicious-software sample local clustering module and a malicious-software sample local analysis module. The invention has the advantages of high analysis accuracy, strong analysis performance and good expandability.

The present invention discloses a load reliability assessment method considering load grading and energy scheduling. The method comprises the following steps: 1. obtaining the original data of a power distribution system and the original meteorological data within a period of time; 2. establishing a reliability model for a distribution type power supply and energy storage combined power generation system; 3. making the local network formed by the distribution power supply and the energy storage an equivalent node recorded as a PCC node; and using the GO model to calculate the equivalent state model and the parameters of the external components relative to the PCC node; 4. classifying the loads in the PCC node, setting the weight coefficients; and using an improved particle swarm optimization algorithm to solve the optimal scheduling; and 5. using Monte Carlo simulation method to calculate the load reliability index. The method of the invention can calculate the reliability of the loads with different characteristics more accurately, and can propose an energy scheduling scheme able to improve the load reliability through the optimization algorithm.

The invention discloses a network topology building method based on end-to-end measurement. An end-to-end network measurement model is built, and a resource transmitting node set and a receiving node set are determined; a measurement message on a leaf node is collected, and collected data is cleaned; a transmission delay time of the end-to-end measurement message is calculated, and the degree of the leaf node in the tree network topology structure is calculated according to TTL fields; the end-to-end delay time similarity calculation formula S=Idk-djI is used to build the network topology. Because the end-to-end network measurement technology is used, judging whether brotherhood fraternity exists between end to end according to the brotherhood leaf node end-to-end delay time similarity can effectively solve the problems of measurement and calculation of the network topology structure under the conditions that network paths and device performance are favourable and no message receiving partial order exists.

The invention discloses a multi-dimensional electric energy meter on-site state test method based on a decision-making tree mass. The method comprises the following steps of building a decision-makingtree by using collected data related to an electric energy meter and provided by an electricity information collection system; forming next-level branches through classifying each decision-making tree; generating the completely grown primary decision-making tree; pruning and optimizing on the primary decision-making tree based on a cost complexity pruning algorithm; using an independent test setto assess the accuracy of the primary decision-making tree after pruning and optimizing; selecting the optimal decision-making tree from each set of secondary decision-making trees, and forming a decision-making tree abnormity diagnostic model by all the optimal decision-making trees, wherein a judgment result wins in a voting manner; and transmitting an operational rule of the decision-making tree abnormity diagnostic model to an electricity information collection system Hadoop big data processing cluster, thus implementing abnormity detection on the abnormal running state of the running electric energy meter.

The invention discloses an infrared radiation spectral characteristic simulation analysis method of a movable target. The method comprises the following steps: firstly, carrying out three-dimensional geometric modeling on the movable target, and dividing the movable target according to regions; establishing a target temperature distribution model, and calculating the temperature of each point of the surface of the target under different observation angles; establishing an infrared atmospheric transmission model, and calculating atmospheric transmittance and atmospheric path radiation; setting parameters of a measurement system and the movable target; calculating radiation energy of an image space of the movable target by utilizing the established target temperature distribution model and infrared atmospheric transmission model; and finally, calculating infrared radiation energy of a point target and a plane target respectively, and drawing a corresponding radiation spectral curve. The method disclosed by the invention is simple, and considered factors are relatively comprehensive; and data can be provided for subsequent detection and tracking and identification of the target.

The embodiment of the invention discloses a calculating method and system for a relationship between the robot position and a planning path. The calculating method for the relationship between the robot position and the planning path comprises the steps that the current position of the robot, the panning path and a target linear segment approaching the planning path are obtained; on the target linear segment, a target coordinate point is determined, wherein the distance from the target coordinate point to the current position meets the first preset distance requirement; according to the position of the target coordinate point on the target linear segment, and a first target terminal and a second target terminal of the target linear segment, an initial coordinate point is determined in a closed interval constituted by the first target terminal and the second target terminal on the planning path; and by taking the initial coordinate point as an initial value, the distance between the current position and the planning path and the projection point of the current position on the planning path are calculated. According to the calculating method and system, the problem of projection point calculation on the real-time planning path is solved, and the real-time and high-precision requirements of robot control are met.

The invention discloses a power distribution Internet of Things CPS management and control method and system based on cloud-side cooperation, and the method comprises the following steps: dividing a power distribution Internet of Things CPS system based on cloud-side cooperation, and obtaining a corresponding centralized-distributed combined control network architecture; carrying out modeling analysis based on a centralized-distributed combined control network architecture to obtain a mathematical model of a power distribution Internet of Things CPS management and control system; based on a mathematical model of a power distribution Internet of Things CPS management and control system, establishing a unit-level CPS mathematical model by taking an edge calculation group as a basic unit; and combining the plurality of unit-level CPS mathematical models in cooperative operation to form a CPS unified calculation model. According to the method, the CPS is reasonably divided by using a cloud-side collaborative centralized-distributed combined control computing system, and a layered collaborative autonomous network architecture is established, so that the computing problem in the CPS can be effectively solved, and efficient utilization of electric energy resources and rapid iteration of services are realized.