134results about How to "Reduce computational efficiency" patented technology

The invention provides a remote sensing image scene classification method based on multi-scale depth feature fusion and transfer learning in order to solve the classification problem of remote sensingimage scenes. The method comprises the following steps: firstly, obtaining a multi-scale remote sensing image by using a Gaussian pyramid algorithm, inputting the multi-scale remote sensing image into a full convolutional neural network, and extracting multi-scale depth local features; cutting the image to a fixed size required by the CNN; obtaining global features of a full connection layer in an input network, using compact bilinear pooling operation to encode multi-scale depth local features and global features obtained by a CNN, fusing the two depth features to jointly represent a remotesensing image, enhancing the mutual relation between the features, and enabling the obtained features to be more distinctive; and finally, classifying the remote sensing image scenes by utilizing a transfer learning technology and combining the two methods. According to the convolutional neural network, a VGG16-Net is used as a basic network.

The present invention relates to a method for the automatic identification of at least one informative data filter from a data set that can be used to identify at least one relevant data subset against a target feature for subsequent hypothesis generation, model building and model testing. The present invention describes methods, and an initial implementation, for efficiently linking relevant data both within and across multiple domains and identifying informative statistical relationships across this data that can be integrated into agent-based models. The relationships, encoded by the agents, can then drive emergent behavior across the global system that is described in the integrated data environment.

The invention relates to a mobile phone information based section speed calculation method. The method is used for calculating the average speed information of a vehicle at each section through the analysis and the treatment of mobile phone information and the combination of traffic road position and direction. The key points of calculating the section average speed lies in massive data processing and precise positioning of a mobile phone user in a movement process. The problems are solved through two-stage three-time map matching in the invention. The map matching at the first stage comprises the following step of: matching Cell with sections, and the map matching at the second stage comprises the following steps of: establishing a subordination relationship of mobile phone data points and the sections through matching user data points with the sections two times, finally calculating the traffic speed of each user on the sections by using the Cell position in which continuous data points are located and the time, removing part of abnormal speeds in the traffic speeds, summing the traffic speeds and averaging the sum to obtain the average speed of each road. In the invention, the large amount of calculation caused by directly matching each data point with the sections is avoided, and the calculation efficiency is improved. The accuracy of the mobile phone data and the sections is improved through a proximity principle and a direction consistency principle.

The invention discloses a self-adaptive control method for diagonal gait of a four-footed robot, which is used for self-adaptively controlling the walking of the four-footed robot in the nonstructural terrain with the diagonal gait. The self-adaptive control method bases on a fuzzy reasoning learning method and a foot trajectory real-time correction method and takes the diagonal gait as the motion mode so as to comprehensively control a robot body to adapt for the nonstructural environment. The gait planning is carried out on the environmental information collected by the sensor through the fuzzy reasoning learning method; the stability judgment is carried out on the gait planning information; if the planned four-footed robot has stable diagonal gait, the joint corner information is sent to a control system; if not, the supporting time of the diagonal feet is corrected by the foot trajectory real-time correction method so that the four-footed robot has better stability when walking under the changed gait in the nonstructural terrain. The self-adaptive control method is used for correcting foot trajectory planning and walking stability under changed gait when the four-footed robot walks diagonally in a dynamic way, thus realizing the self-perception, self-correction and self-adjustment of the four-footed robot.

The invention discloses a visual configuration method for index calculation. The visual configuration method for the index calculation includes following steps: S01, building models of business objects; S02, building a model of basic data: selecting attributes of the business objects in visualization mode from the models of the business objects, built in the step S01, in an FLEX environment, and connecting the business objects through operational characters so as to form a calculation link of basic data items; S03, building a model of the index calculation: resolving logic of the index calculation into the basic data items, weighting coefficients and the operational characters, and forming a topological graph model of the logic of the index calculation in the FLEX environment; S04, calculating an index: using numerical value of the basic data items and an algorithm described by the operational characters to calculate layer by layer so as to push deep into a root node, and finally obtaining an index value on the root node. The visual configuration method for the index calculation organically integrates technologies of concurrent calculation, memory calculation, visual index calculation model configuration together, and develops the advantages of being parallel in clustering of the concurrent calculation, being high in efficiency of the memory calculation and the like.

The invention relates to a spot-welding connection failure value simulation system in the car bodywork manufacturing engineering field. In the invention, a circle-stress failure criterion generating module obtains nugget circle stress distribution form according to the failure mode of the spot-welding connection in a strike process to generate the circle stress failure criterion; a failure extreme value stress and an index-parameter obtaining module obtain welding-spot critical failure stress extreme values and dimensionless stress sub-item index initial values under the condition of the peak values of four exterior loads; four types of sample value models which are pulling-cutting, pulling-extending, pulling-bending and pulling-twisting are established in a welding-spot connection value simulation module; an explicit dynamics analysis module couples circle stress failure criteria and obtains the continuous distributed inner stress of a welding-spot unit model and an extreme value stress borne by the welding-spot unit during failure time by simulation; a welding-spot failure material databank module stores the critical failure stress extreme values and the dimensionless stress sub-item index initial values of each stress component. The system not only can enhance the impact simulation precision without impact on the calculation efficiency, but also can reduce the design and exploitation cost.

The invention refers to a method for simultaneous location and map creation. The method can be used for automatic positioning of a self-moving device, especially for a self-moving ground dust collecting device. The device is preferentially provided with a plurality of moving wheels driven by a motor, wherein the device is provided with an obstacle recognition device composed of an optical emitting and receiving unit, wherein a plurality of possible positions and directions (mass point 6) are further calculated from a beforehand calculated an assumed first position; and according to a second position occupied by the device after movement, by means of measurement results of the optical unit, one of the mass points is assumed as a new position of the device according to a predetermined selection algorithm considering probability. In order to further improve the method of this type, it is suggested that the calculation of the mass point is limited from the aspects of a solid angle and a distance range determined by a predetermined moving direction and moving distance or a moving speed.

The invention belongs to the five-axis numerical control machine tool machining field, and provides a semi-analysis modeling method of a contact area between a ball-end milling cutter and a guide vane. The semi-analysis modeling method of a contact area between a ball-end milling cutter and a guide vane includes the following steps: performing cutter path programming, and setting numerical control machining parameters; establishing a workpiece coordinate system, and acquiring the parameter equations of the design surface and the machining surface of the vane, and the movement track equation of the ball center of the milling cutter; determining the cutter contact point and the cutter location point during the machining process, taking the cutter contact point and the cutter location point as the origin to respectively establish a three dimensional rectangular coordinate system; converting solving problems of No.1-No.4 boundary curves of the contact area into projection equation solving problems of the curves in an XL-YL two dimensional coordinate system, and respectively solving the projection equations of the No.1-No.4 boundary curves; and obtaining the intersection points of the No.1-No.4 boundary curves, by mutually making a simultaneous system of equations, and obtaining the projection of the contact area. The semi-analysis modeling method of a contact area between a ball-end milling cutter and a guide vane can solve the problem of performing analysis modeling of the contact area during the process of processing a straight-vane-body vane molded in a vane grid by means of a ball-end milling cutter on a five-axis numerical control machine tool.

The method generates a simulated object represented by finite elements each comprising two or more nodes that represent a first physical object, the first simulated object comprising a plurality of segments placed adjacent to each other to form a surface of the first simulated object. The method further includes generating a second simulated object. The method determines the distance between individual segments of the first simulated object and the plurality of segments of the second simulated object. The method determines a stiffness matrix and force vectors for the at least one segment of the first simulated object that is in contact with at least one segment of the second simulated object. The method transforms the stiffness matrix and the force vector from the segments to determine a stiffness matrix and a force vector on the two or more nodes of the finite element representation of the physical objects.

The invention discloses a polarimetric SAR (synthetic aperture radar) image target detection method based on multipolarization features and an FCN (fully convolutional network)-CRF (conditional random field) fusion network. The invention aims to solve the problem of low detection accuracy of a polarimetric SAR artificial target of the prior art. The method includes the following steps that: a polarized SAR image to be detected is inputted, and Lee filtering is performed on the polarization coherent matrix T of the polarimetric SAR image; Pauli decomposition is performed on a polarimetric scattering matrix S, so that a pixel point-based feature matrix F1 can be formed; and Yamaguchi decomposition is performed on the filtered coherent matrix T, so that a pixel-based feature matrix F2 can be formed. According to the method of the invention, the multi-polarization feature and the FCN-CRF-based fusion network are applied to the detection of a polarimetric SAR artificial target, and therefore, the detection accuracy of the polarimetric SAR artificial target can be improved; and the method can be applied to multi-target classification tasks.

The invention discloses a decentralized access control method for safe sharing of data in a smart power grid. The method comprises the following main implementation steps: 1, constructing an access control system; 2, initializing an access control system, and generating a global public parameter GP of the system; 3, initializing an authorization mechanism; 4, generating a user key; 5, generating acloud server private key, a signature public key and a signature private key; 6, generating a final ciphertext; 7, verifying the identity of the user and decrypting the message; and 8, deleting the user to be revoked from the private key list of the cloud server. According to the method, multiple mechanisms are adopted to jointly undertake tasks of the system, the efficiency of the system is improved, a large amount of encryption and decryption calculation is carried out by the cloud server, the calculation overhead of a data user is saved, and zero knowledge is added to prove that the identity information of the user is protected in the interaction process of the user and multiple authorization mechanisms.

The invention discloses a pedestrian detection method based on wavelet fractal characteristic, which relates to the technology of intelligent vehicles in an intelligent traffic system. The pedestrian detection method comprises the following steps: (1) reading a training sample set and standardizing all sample images to the pixel size of 48*96; (2) carrying three times of two-dimensional wavelet transformation on the sample images, and getting six wavelet sub-graphs in the second layer and the third layer; (3) getting absolute values of the wavelet sub-graphs obtained in the step (2), and carrying out stretching and scaling so that the numeric area is uniformly mapped to 0 to 255; (4) respectively getting fractal dimension vectors of the wavelet sub-graphs treated in the step (3); (5) standardizing the fractal vectors corresponding to each sub-graph obtained in the step (4); (6) combining the fractal vectors standardized in the step (5), and obtaining the wavelet fractal characteristic vector of 6*(n-1) dimension; and (7) using the wavelet fractal characteristic obtained in the step (6) for training a soft-supporter vector machine, and using the obtained discriminant function for realizing pedestrian detection. The invention has the characteristics of concise characteristic expression form, strong characteristic distinguishing capability and higher pedestrian detection efficiency.

The invention discloses a pre-stack seismic inversion method suitable for shale reservoirs with strong horizontal stratification structures and the like. Different from a conventional inversion methodbased on an equivalent VTI medium reflection coefficient formula, the method selects a recursive matrix algorithm as a forward modeling drive of prestack inversion. According to the recursive matrixalgorithm, a wave equation accurate solution and a recursive data mode are combined, and the internal complex wave field effect of shale and the like with a strong stratified structure can be effectively simulated. The recursive matrix based prestack inversion method is based on isotropic hypothesis, can reduce target parameters and improve inversion stability, and can be used for estimating elastic parameters of stratified structure reservoirs such as shale, eliminating the influence of a complex wave field in the shale on an underlying stratum and recovering attribute characteristics of theunderlying stratum. A synthetic record test of model parameters proves the effectiveness of the method provided by the invention.

The invention discloses an anemometer tower microcosmic site selection method based on a wind acceleration factor, and belongs to the technical field of wind energy resource evaluation. The method comprises the following steps: firstly, obtaining a digital topographic map of a to-be-developed wind power plant area, carrying out mesh generation on the digital topographic map, and then carrying outdirectional calculation based on computational fluid dynamics to obtain wind acceleration factor distribution data in different wind direction sectors; then determining a plurality of alternative anemometer tower setting point locations by combining the main wind direction information and the wind acceleration factor distribution data; and finally, screening and rechecking the alternative anemometer tower setting point positions, and finally determining the anemometer tower setting point positions meeting the anemometer tower conditions. A reliable basis at the same stage and data collection degree is provided for microcosmic site selection of the anemometer tower. Therefore, the accuracy of site selection is remarkably improved, the situations that resource measuring points are not ideal,wind power plant development decisions are misled and the like possibly caused by the fact that a high-point anemometer tower is blindly selected according to experience are effectively avoided, andthe development time and investment cost of the wind power plant in complex terrains are greatly reduced.

The invention discloses a method of analyzing structural reliability of an aeroengine turbine disc; a random variable mathematical model is established according to non-uniformity of turbine disc material, and a stress analysis model is established according to stress condition; PDS (probabilistic design system) in ANSYS environment is used to perform Monte Carlo sampling on random variables of a turbine disc three-dimensional model to obtain maximum stress MAXSTRESS distribution of the turbine disc under different conditions; Monte Carlo simulation tests of different levels are performed on the random variables on MATLAB (matrix laboratory) platform, and reliability of the turbine disc is calculated according to structural functions to obtain the structural reliability of the turbine disc under different conditions; sensitivity of the random variables influencing the structural reliability of the turbine disc is calculated according to sensitivity equation. The method has the advantages of high convergence speed, low cost and high precision, and theoretical basis is provided for the optimal design and reliability management for a turbine disc.

The invention discloses a frequency domain multi-conductor transmission line electromagnetic pulse response rapid modeling method based on waveform relaxation iteration. A frequency domain calculationmodel for solving coupling of the electromagnetic pulse to the multi-conductor transmission line based on a waveform relaxation algorithm and analytical iteration is established in the frequency domain; A combined voltage wave equation and a BLT supermatrix equation are used; the mutual coupling influence between the transmission lines is equivalent to virtual excitation sources which are continuously distributed on the transmission lines; An analytical expression of voltage and current responses along a transmission line is deduced and given, large-scale matrix inverse operation is convertedinto a series of iteration processes, meanwhile, each iteration step can give an analytical solution, time-consuming numerical integration is avoided, and the calculation efficiency is improved.

An embodiment of the invention provides three-dimensional fully convolutional network realizing equipment The equipment comprises the components of a receiving unit which is used for receiving initialpoint cloud information that has a tree-shaped storage structure; and a convolutional unit which is used for performing multilayer convolutional and fully convolutional operation on the initial pointcloud information through the three-dimensional fully convolutional network, thereby obtaining reverse volume pixel prediction information which comprises a plurality of kinds of information. Becausea 3D environment comprises abundant set characteristic information, a traditional vision-based settlement method cannot effectively perform real-time analysis on a 3D scene, the embodiment of the invention provides a 3D convolutional solution based on the tree-shaped structure, semantic classification to the 3D point cloud environment can be finished, thereby improving understanding of the 3D environment.

The invention discloses an improved exponential time integral constructing method in a complicated dispersive medium, and is applied to the field of three-dimensional electromagnetic analysis numerical values. The improved exponential time integral constructing method in the complicated dispersive medium comprises the following steps: carrying out three-dimensional molding on a to-be-analyzed multi-scale target which comprises the complicated dispersive medium, and establishing a corresponding geometry structure model; splitting the established geometry structure model by using a tetrahedral mesh; separately integrating all local field components into respective corresponding global unknown quantities to construct a global semi-discrete format at first; and then integrating electric fieldand magnetic field components into an unknown quantity to obtain an ordinary differential equation, namely an electric field and magnetic field updating equation. Therefore, updating of electric fields and magnetic fields only needs vector manipulation, the circumstance that matrix exponentials are introduced to implement updating by a traditional exponential time integral method is avoided, to-be-solved matrix exponent dimensionality is reduced remarkably, the matrix exponent calculating efficiency is reduced effectively, and therefore, the simulating efficiency based on a time domain discontinuous Galerkin method of the exponential time integral method is improved.

The invention provides a precise processing method and system for a rotary large-breadth optical satellite. The precise processing method for the rotary large-breadth optical satellite comprises the steps of: acquiring optical camera area array subsystem parameters and observation data; according to the optical camera area array subsystem parameters and the observation data, constructing a geometric transformation relationship between a rotary optical satellite body and an optical camera area array subsystem focal plane to obtain a satellite body and camera focal plane transformation matrix; calculating rotary optical satellite attitude and orbit control data at the imaging moment; constructing a rotary large-breadth optical satellite precise processing geometric model according to the satellite body and camera focal plane transformation matrix and the rotary optical satellite attitude and orbit control data; constructing a rational function fitting model of a standard image product according to the rotary large-breadth optical satellite precise processing geometric model; and carrying out rational polynomial coefficient parameter calculation according to the rational function fitting model of the standard image product.

The invention discloses an intelligent transportation scheduling management method based on a neural network genetic algorithm. The method comprises the following steps: A1, establishing a data moduleaccording to a transportation scheduling management system; a2, generating an initial population G, wherein the initial population comprises N chromosomes, and the chromosome encodes the material M (Ma and/or Mb) according to the characteristics selected as required according to any length; constructing a neural network, wherein the delivery point is a neural network input node, the receiving point is an output node, and the transfer station is a hidden layer; a4, starting iteration. A neural network algorithm is added into the genetic algorithm, so that the defects of low calculation efficiency, easiness in falling into local optimum, difficulty in convergence and the like in the genetic algorithm are avoided, and premature convergence or a large number of iterative complex calculationsare avoided. The fitness evaluation function evaluates from two opposite angles of economic applicability e and time length t, and the two opposite functions are balanced with each other, so that a rapid and economic logistics scheduling scheme is realized.