114 results about "Boundary model" patented technology

Embodiments of systems and methods for tuning a turbine are provided. In one embodiment, a method may include receiving at least one of a measured operating parameter or a modeled operating parameter of a turbine during operation; and tuning the turbine during operation. The turbine may be tuned during operation by applying the measured operating parameter or modeled operating parameter or parameters to at least one operational boundary model, applying the measured operating parameter or modeled operating parameter or parameters to at least one scheduling algorithm, comparing the output of the operational boundary model or models to the at output of the scheduling algorithm or algorithms to determine at least one error term, and closing loop on the one error term or terms by adjusting at least one turbine control effector during operation of the turbine.

The prevention of vehicle accidents is targeted. A road texture model is created based on a vehicle camera image. An initial vehicle location estimate is determined, and map imagery is obtained based on this location estimate. A refined vehicle location is determined using visual egomotion. In particular, 3D features of the vehicle image and the retrieved map imagery are identified and aligned. A map image is selected based on this alignment, and the location associated with the map image is modified by a displacement between the selected map image and the vehicle image to produce a refined vehicle location. A road boundary model is created based on the road texture model and the refined vehicle location, and a road departure model is created based on the road boundary model and vehicle odometry information. The operator of the vehicle is warned of a road departure based on the road departure model.

A method for generating a prosody model that predicts prosodic parameters is disclosed. Upon receiving text annotated with acoustic features, the method comprises generating first classification and regression trees (CARTs) that predict durations and F0 from text by generating initial boundary labels by considering pauses, generating initial accent labels by applying a simple rule on text-derived features only, adding the predicted accent and boundary labels to feature vectors, and using the feature vectors to generate the first CARTs. The first CARTs are used to predict accent and boundary labels. Next, the first CARTs are used to generate second CARTs that predict durations and F0 from text and acoustic features by using lengthened accented syllables and phrase-final syllables, refining accent and boundary models simultaneously, comparing actual and predicted duration of a whole prosodic phrase to normalize speaking rate, and generating the second CARTs that predict the normalized speaking rate.

The invention discloses a dynamic gesture sequence real-time recognition method, a dynamic gesture sequence real-time recognition system and a dynamic gesture sequence real-time recognition device. The dynamic gesture sequence real-time recognition method comprises the steps of: separately acquiring a color image and a depth image containing an object to be recognized; performing human body regiondetection and segmentation according to the acquired color image and depth image, so as to obtain a human body region; detecting and segmenting hand regions in the human body region so as to obtain the hand regions; adopting a skin color mode with illumination invariance and an elliptical boundary model based on Gaussian distribution for tracking hands dynamically according to the hand regions; adopting a method based on gesture trajectory and static attitude matching for performing space-time gesture sequence detection according to dynamic tracking results of the hands, so as to obtain a dynamic gesture sequence; and modeling and classifying the dynamic gesture sequence. The dynamic gesture sequence real-time recognition method, the dynamic gesture sequence real-time recognition system and the dynamic gesture sequence real-time recognition device improve the robustness of gesture recognition through utilizing the depth information and adopting the skin color mode with illumination invariance and the elliptical boundary model based on Gaussian distribution, have good recognition effect, and can be widely applied in the fields of artificial intelligence and computer vision.

Systems and methods for training semantic segmentation. Embodiments of the present invention include predicting semantic labeling of each pixel in each of at least one training image using a semantic segmentation model. Further included is predicting semantic boundaries at boundary pixels of objects in the at least one training image using a semantic boundary model concurrently with predicting the semantic labeling. Also included is propagating sparse labels to every pixel in the at least one training image using the predicted semantic boundaries. Additionally, the embodiments include optimizing a loss function according the predicted semantic labeling and the propagated sparse labels to concurrently train the semantic segmentation model and the semantic boundary model to accurately and efficiently generate a learned semantic segmentation model from sparsely annotated training images.

The invention relates to the field of data processing, in particular to a wind power modeling and performance evaluating method based on a confidence equivalent power curve band. The wind power modeling and performance evaluating method includes the steps that preliminary screening and rejection are conducted on an abnormal data sample; wind speed is divided into three regions, and a kernel density estimation method is used for obtaining wind speed and power probability distributions in each region through statistics to obtain the Copula function in each region; the maximum likelihood estimation method is adopted to obtain confidence equivalent power boundary model in the corresponding region; a piecewise cubic Hermite interpolating polynomial method is used for reconstructing missing datato complete original data sample cleaning; the average value of the confidence bandwidth ratio is used as the model performance evaluation index, a d-fold cross validation method is used for validating the upper and lower boundary models in different regions, when the index is basically stabilized at a certain constant value, the up and lower boundary models of the different regions are determined; a sliding time window method is adopted to updating data, the deviation degree of the confidence bandwidth ratio is used as a triggering condition, and the up and lower boundary models are updatedwhen the deviation degree exceeds a certain threshold value.

The invention provides a sand shale rock reservoir stratum pressure earthquake prediction method and device; the method comprises the following steps: calculating the elastic modulus, the microfissure density and microfissure closed rate factors under the limited effective stress according to a Hashin-Shtrikman boundary model and the porosity and shale content of an object sand shale rock reservoir in a research zone; determining the relation between the dry rock elastic modulus and effective stress according to the elastic modulus, the microfissure density and microfissure closed rate factors; determining the effective stress coefficient of the sand shale rock along with the depth changes according to said relation; predicting the stratum pressure of a single well to-be-detected object layer according to said relation and the effective stress coefficient; using an elastic modulus data body obtained by inversion and taking the single well stratum pressure as constraints, and carrying out three dimensional earthquake stratum pressure prediction on the research zone according to said relation and the effective stress coefficient. The method and device can improve the stratum pressure prediction accuracy.

The invention discloses an adaptive road aerial view transformation method and a road lane detection method. The transformation method comprises the steps: road boundary feature extraction, road boundary segmentation extraction, road boundary point extraction, road boundary modeling and matching, the obtaining of an aerial view transformation matrix through the transformation of adaptive aerial view transformation, and the mapping of a road image gray scale map to an aerial view map. The detection method comprises the steps: firstly carrying out the marking and filtering of lanes, and carrying out feature extraction and the estimation of the positions of the lane marks; and secondly extracting the positions of the lane marks in a road scene image through the positions of the lane marks The methods are better in robustness, and be used for processing a road surface comprising a large amount of shadow interference, solve a problem that there are fixed transformation parameters in a conventional method, can automatically adjust the parameters of a system, and extract the positions of lane mark lines more accurately.

The invention discloses a deep learning based time sequence motion identification method. The method comprises video feature extraction and construction of a time boundary regression model. Aimed at the problem that that a long motion feature cannot be expressed effectively enough in the boundary detection process, inter-frame information and intra-frame information are extracted simultaneously via a double flow network to obtain a feature sequence of a video unit, a multi-scale short motion segment extracting scheme combining context information is provided, the accuracy of subsequent regression is improved effectively, the time boundary model is trained via a feature sequence, training time of the model is reduced, and the calculation efficiency is improved. The improved time boundary regression model including an improved multi-task multilayer sensor and a brand new splicing mechanism aimed at long motions is provided aimed at the problem of inaccurate boundary regression of the long motion; and on the basis that the motion type is accurate, the accuracy of time boundary regression of the long motion is improved effectively, the predicted motion segment is more overlapped with the practical motion segment, and the identification rate of the time sequence motions is improved.

The invention provides a method for modeling and simulating a three-dimensional discrete element for shale pneumatic and hydraulic fracturing. The method includes the steps of (1) establishing a three-dimensional close packing model with specific elastic property and destructive property according to a conversion formula, and establishing an integrated isotropic random module through particle diameter normal distribution adjustment and random packing, (2) adjusting the diameter of particles according to given confining pressure, setting particle anisotropic ellipsoid parameters, moving the boundary of the models, adjusting the confining pressure to the given value, and finally connecting a particle solidification model to obtain an anisotropic shale model with specific confining pressure and mechanical property, (3) establishing a hydraulic pressure cavity model and a hydraulic pressure boundary model according to requirements, generating hydraulic pressure and stress waves through mutual extrusion of water particles, and simulating generation and development of hydraulic fractures. The method avoids human influences of a traditional method based on continuity mechanics, can simulate the processes of hydraulic pressure propagation and hydraulic fracturing more really, and achieves predication of shale pneumatic and hydraulic fracturing.

A method for generating a prosody model that predicts prosodic parameters is disclosed. Upon receiving text annotated with acoustic features, the method comprises generating first classification and regression trees (CARTs) that predict durations and F0 from text by generating initial boundary labels by considering pauses, generating initial accent labels by applying a simple rule on text-derived features only, adding the predicted accent and boundary labels to feature vectors, and using the feature vectors to generate the first CARTs. The first CARTs are used to predict accent and boundary labels. Next, the first CARTs are used to generate second CARTs that predict durations and F0 from text and acoustic features by using lengthened accented syllables and phrase-final syllables, refining accent and boundary models simultaneously, comparing actual and predicted duration of a whole prosodic phrase to normalize speaking rate, and generating the second CARTs that predict the normalized speaking rate.

The invention relates to a universal method of boundary modeling based on distinct element method, which is applicable to analyze and evaluate the service performance of a mechanical part through distinct element method. The technical solution comprises the following steps: step 1, alternately identifying a part surface for contacting particle material in the mechanical part; step 2, automatically identifying that the part surface is a regularly curved surface or an irregularly curved surface; step 3, automatically identifying the missing part of the regularly curved surface; step 4, dispersing the irregularly curved surface into combination of regularly curved surfaces; step 5, reading the parameter of the regularly curved surface; step 6, alternately adding the motion parameters and material characteristic parameters of the curved surface; step 7, storing boundary model data. By changing the CAD (Computer - Aided Design) model of the mechanical part, the performances of mechanical parts with different structures and sizes can be analyzed so as to optimize the structures and sizes of the mechanical parts, thus solving the technical problem on the optimum design of the mechanical parts and bringing about remarkable technical effects.

A construction method of a flow field numerical model in a minimum quantity lubrication cutting area comprises the following steps of building a cutting two-dimensional or three-dimensional model; conducting grid dividing on the module; leading the grid-dividing-completed-model in simulation software and using a steady state flow field computing method to conduct simulation after defining entry conditions and exit conditions; building a motion model and a cutting heat model and converting the motion model and the cutting heat model into recognizable language of the simulation software; selecting an appropriate nozzle model and a wall film boundary model; leading the models in a steady state flow field and using an unsteady state computing method to conduct simulation to obtain velocity field distribution, pressure field distribution, temperature field distribution, fogdrop diameter and fogdrop speed. The construction method can have flow field influences of minimum quantity lubrication on the cutting area during cutting and influences on the minimum quantity lubrication on workpieces and temperature fields inside a tool, provides basis for minimum quantity lubrication system setting, and simultaneously is used for parameter optimization of a minimum quantity lubrication system.

The invention discloses a modeling method of the boundary of a curve type farmland operation area. The modeling method comprises the steps that key points of the boundary of the curve type farmland operation area are collected, the point with the maximum abscissa value and the point with the minimum abscissa value of all the key points are found, and an upper boundary model and a lower boundary model are established through the cubic spline function. According to the modeling method of the boundary of the curve type farmland operation area, the boundary of the curve type farmland operation area and the peaks of obstacles in the farmland operation area can be accurately recorded and drawn, so that the accurate farmland operation area is provided for operators of an agricultural plant protection machine, and a reliable basis is provided for the route and path planning of the agricultural plant protection machine and calculation of the farmland operation area.

The invention discloses a clothing simulation method and a system thereof. The method comprises the following steps: S1, inputting parameters to a simulation system through a parameter definition module and solving the same, wherein the input parameters comprise a physical parameter of a clothing material, a hot human physiological parameter, an environmental climate parameter and a human activity behavior parameter; S2, building a human body-clothing-environment stimulation system, including an environmental climate boundary model, a multimodal mathematical model of hot human physiological balance adjustment and a mathematical model of clothing and environment heat and moisture transfer and exchange, and building a digital simulation equation of a clothing environment and a digital simulation equation of hot human physiological balance adjustment respectively according to the mathematical model of clothing and environment heat and moisture transfer and exchange and the multimodal mathematical model of hot human physiological balance adjustment; and building a multimodal modularized calculation model in the stimulation system; S3, obtaining simulation data, and outputting a 3D graphics showing clothing performance through an image output module.

The invention discloses a lightweight processing method for a product model. The method comprises the steps of analyzing the product model and acquiring a characteristic set of the product model; according to a preset lightweight policy, determining deletable characteristics in the characteristic set; and deleting the deletable characteristics, reserving a geometric structure characteristic and a size characteristic of the product model, and generating a lightweight product model with the same shape structure as the product model. According to the lightweight processing method and apparatus for the product model, the characteristic removal processing can be performed, only a final boundary model structure is reserved, and the assembly level fusion processing can be performed, so that the model storage space can be greatly saved and the model processing speed can be greatly increased; and the lightweight processing of parts is automatically finished, so that the model can be assembled in a lightweight form, the complicated characteristics and the characteristic regeneration problem are avoided, and the working efficiency of designers is improved.

The present invention relates to a method for designing a highly uniform magnetic field coil, in particular to a method for designing a highly uniform magnetic field coil in a magnetic shielding environment. According to the method provided by the present invention, a ratio of a magnetic field calculated by the Biot-Savart Law under a condition of a free boundary to a magnetic field calculated under a condition of establishing a magnetic shielding boundary model by a finite element method is set as an equivalent gain coefficient of amplification of a magnetic shielding boundary to a coil magnetic field coefficient; a field and circuit integrated method is adopted; and coil parameters are subjected to optimal design by using the minimum change rate of the coil magnetic field as the target. Compared with an existing coil design method under the magnetic shielding boundary, which completely adopts the finite element method to carry out design, the method disclosed by the present invention can obtain a coil structural parameter with a higher uniformity and has high computation efficiency and feasibility.

The invention relates to a provincial dispatching and local dispatching-integrated joint risk assessment method. The method includes the following steps that: information such as a provincial dispatching and local dispatching integration model and measurement is read in; a preconceived fault set is generated automatically; provincial dispatching carries out preconceived fault scanning on a whole grid; local dispatching load transfer plans are optimized; the provincial dispatching carries out risk assessment on the load-transferred grid; preconceived fault load switching is carried out based on sensitivity; and a provincial dispatching and local dispatching-integrated risk assessment result is outputted, and early warning is carried out. With the method of the invention adopted, provincial dispatching and local dispatching joint risk assessment is realized, and the comprehensiveness of risk assessment is realized. According to the method, the method is realized based on a provincial and local integration platform, links such as provincial and local boundary model analysis and splicing are eliminated, the consistency of provincial dispatching risk calculation basic data and local dispatching risk calculation basic data can be ensured; and a sensitivity-based algorithm is adopted, so that the result of risk assessment can more accord with the actual situation of the power grid, so that dispatchers can accurately locate faults and timely take corresponding control measures, and therefore, the safe operation level of the power grid can be improved.

The prevention of vehicle accidents is targeted. A road texture model is created based on a vehicle camera image. An initial vehicle location estimate is determined, and map imagery is obtained based on this location estimate. A refined vehicle location is determined using visual egomotion. In particular, 3D features of the vehicle image and the retrieved map imagery are identified and aligned. A map image is selected based on this alignment, and the location associated with the map image is modified by a displacement between the selected map image and the vehicle image to produce a refined vehicle location. A road boundary model is created based on the road texture model and the refined vehicle location, and a road departure model is created based on the road boundary model and vehicle odometry information. The operator of the vehicle is warned of a road departure based on the road departure model.

Embodiments of systems and methods for tuning a turbine are provided. In one embodiment, a method may include receiving at least one of a measured operating parameter or a modeled operating parameter of a turbine during operation; and tuning the turbine during operation. The turbine may be tuned during operation by applying the measured operating parameter or modeled operating parameter or parameters to at least one operational boundary model, applying the measured operating parameter or modeled operating parameter or parameters to at least one scheduling algorithm, comparing the output of the operational boundary model or models to the output of the scheduling algorithm or algorithms to determine at least one error term, and closing loop on the one error term or terms by adjusting at least one turbine control effector during operation of the turbine.

The invention belongs to a target boundary extraction method based on laser stripe geometric feature mutation, which belongs to the field of visual measurement. The method comprises steps: a mode of combining laser and binocular vision is firstly adopted to acquire a laser stripe image; according to a boundary laser stripe condition existing in an actual measurement process, three kinds of boundary models are built; differentiate operation is used to carry out derivation on a boundary eigenvector, and target boundary feature point coordinates through rough extraction are obtained; and according to a moment invariance principle, target boundary feature point coordinates through fine extraction are obtained, and quick and high-precision extraction on the target boundary is realized. The method overcomes phenomena that during the existing boundary extraction process, many noisy points are generated at the boundary of two ends of the laser stripe, and due to influences of a spatial position relationship, the laser stripe is bent, dislocated or broken and disconnected and the gray distribution level and the width presented by the laser stripe are different, and quick and high-precision target boundary extraction is realized.

The invention provides a high-acidity gas field wellbore string material optimization evaluation system and method, and belongs to the field of high-acidity gas field development. The high-acidity gas field wellbore string material optimization evaluation system includes a corrosion rate prediction model and a stress corrosion cracking boundary model; the corrosion rate prediction model includes H2S partial pressure-temperature, H2S partial pressure-CO2 partial pressure and elementary substance S-pH value corrosion rate prediction drawing boards; the stress corrosion cracking boundary model includes a failure boundary of a material under conditions of the temperature and the H2S partial pressure and a failure boundary of the material after sulfur is separated out under the conditions of the temperature and the H2S partial pressure. The system and the method solve the problems that due to a corrosion simulation evaluation method and a corrosion prediction model and software, a gas field highly containing H2S / CO2 is lack of basis on material selection, and provide a theoretical support for high-acidity gas field development and safety production.

The invention provides a calculation and real-time monitoring method for a blast furnace tuyere rotation area boundary, and relates to the technical field of blast furnace ironmaking processes. The method comprises the following steps: firstly, establishing a depth calculation model of a convolution region according to a forming principle of the blast-furnace tuyere convolution region to obtain a calculation formula of the depth of the convolution region and obtain a change rule of the depth of the convolution region; establishing a boundary model of the blast-furnace tuyere convolute area through the depth model of the blast-furnace tuyere convolute area, and determining a calculation formula of the boundary of the convolute area; then obtaining modeling parameters, analyzing the influence of the modeling parameters on the convolution region boundary model, and determining main parameters influencing the convolution region boundary; and finally, calculating the height of the convolution region by using a convolution region boundary calculation formula. And when the height or the depth of the convolution area exceeds a set range, the height or the depth of the convolution area is recovered to a normal range by adjusting the blast air pressure and the blast air volume. The method can monitor the change conditions of the depth and boundary of the convolution area in real time, and provides safety guidance for actual production of the blast furnace.

The invention discloses a tailings pond dry beach parameter monitoring system and method. The system comprises a plurality of laser range finders which are arranged in parallel on the top of a dry beach dam body, a first communication module, an ultrasonic level gauge arranged on the tailings pond area water level, a second communication module and an upper computer. The laser range finders are in communication with the upper computer through the first communication module. The ultrasonic level gauge is in communication with the upper computer through the second communication module. A plurality of dry beach monitoring points are selected freely on a dry beach, each dry beach monitoring point is monitored by a laser range finder, the system is used for monitoring the dry beach, a dry beach water level boundary model is established and corrected in real time through data collected in advance and data collected in real time, the real-time area of the dry beach is obtained, dry beach measuring real-time performance and accuracy are improved, accurate information is provided for early warning, and potential safety hazards are avoided.

A method for calculating parameters about an axisymmetric body in a cavity is provided. The user provides data describing the body, a cavity estimate, and convergence tolerances. Boundary element panels are distributed along the body and the estimated cavity. Matrices are initialized for each panel using disturbance potentials and boundary values. Disturbance potential matrices are formulated for each panel using disturbance potential equations and boundary conditions. The initialized matrices and the formulated matrices are solved for each boundary panel to obtain panel sources, dipoles and cavitation numbers. Forces and velocities are computed giving velocity and drag components. The cavity shape is updated by moving each panel in accordance with the calculated values. The method then tests for convergence against a tolerance, and iterates until convergence is achieved. Upon completion, parameters of interest and the cavity shape are provided. This invention also allows determiniation of cavity shape for a cavitation number.

The invention discloses a method for accurately positioning the iris center of a low-resolution image. The method comprises the following steps of acquiring face feature points, and extracting an eyearea image according to the obtained face feature points; performing binarization treatment on the eye area image and recognizing the eye state, namely eye opening and eye closing recognition; roughlypositioning the iris center of the eye area image through a double-circle activity boundary model; extracting the edge of the iris and accurately positioning the iris center of an eye-opening image by adopting the least binary circle fitting algorithm; and accurately positioning the iris center of an image with less than two edges and the iris center of an eye-closing image by adopting the iris center quality evaluation and binarization communication domain method. The invention also discloses an extraction device for accurately positioning the iris center of the low-resolution image. According to the method, the parameters of the double-circle activity boundary model are initialized by utilizing eye anatomical parameters, and an accurate and practical algorithm for extracting the iris edge is designed. Therefore, the calculation efficiency and the accuracy of iris center positioning are improved.

The invention discloses a transient electromagnetic inversion method for a sharp boundary model. The method comprises the following steps: secondary field data observed via a transient electromagnetic method is input, inversion parameters are input and an inversion initial model is built; a mismatch general function between observed data and forward modeling theory data is built, a minimum gradient support general function used for restraining resistivity change is built, an inversion object function is built via the mismatch general function and the minimum gradient support general function, the initial model is used as initial conditions, a Gauss-Newton method is used for solving resistivity corresponding to a minimum target function, and an inversion result is obtained. Via the inversion method, a resolution of a underground resistivity abrupt change surface, a defect of unclear electrical interface reflection in an Occam inversion algorithm can be compensated, characteristics of stability and simplicity of the Occam inversion algorithm are kept in the transient electromagnetic inversion method, and the transient electromagnetic inversion method is of great significance for transient electromagnetic data inversion and late stage data interpretation.

The invention discloses an ABAQUS three-dimensional infinite element boundary rapid modeling method based on an INP file and belongs to the field of infinite element boundary modeling. According to an infinite unit defined theory, a method defined by a node and a unit in an ABAQUS input file.inp is utilized and an infinite unit defined data row is output with the help of a Python programming technology; and an input file is generated according to a syntax rule of the INP file and is introduced into ABAQUS / CAE, so that automatic generation of an ABAQUS infinite element boundary is realized. The ABAQUS infinite element boundary rapid modeling method is simple to operate and an ABAQUS infinite element boundary model can be generated within short time; and the ABAQUS modeling speed is extremely improved.

The invention provides a ship collision risk determination method based on a digital ship domain model, and the method is characterized in that the method comprises the following steps: 1, building aship domain outer boundary model; 2, constructing a digital ship domain model based on the ship domain outer boundary model; and step 3, based on the digital ship domain model, constructing a collision dynamic risk characterization model among single ships and multiple ships in the water area. By applying the method, the accuracy of ship collision risk early warning and forecasting can be effectively improved, the method has important significance in reducing ship collision accidents, and meanwhile, the reasonable layout of maritime affair safety supervision and search and rescue resources canbe supported and optimized.

The invention provides an equipment fault self-adaptive upper and lower early warning boundary generation method based on a convolutional neural network, which can be used for fault diagnosis of a chemical fiber winding machine in a spinning process. The method comprises a convolutional neural network model for interval prediction and an upper and lower boundary model for interval self-adaptive generation and classification. According to the invention, fault diagnosis is carried out on vibration signals collected by a chemical fiber winding machine during the spinning process, the defects thatan existing fault diagnosis technology is not high in accuracy and is easily influenced by human factors are overcome, a cost-sensitive learning module is introduced to optimize a loss function in the convolutional neural network updating iteration process, and therefore a machine learning fault detection method with the misclassification cost as the optimization target is obtained. The method provided by the invention has good practicability under the condition of unbalanced samples.