59 results about "Model building process" patented technology

The invention discloses a method for constructing a tunnel BIM model based on a Revit platform. According to the method, a tunnel component parameterized family is constructed based on a Revit software platform. A tunnel BIM model is built, classification and summarization are carried out to form a component family library, all component families in the family library are directly called when thetunnel BIM model is built, parameterized automatic driving is achieved through an external data file, parameters of components of the same type can be rapidly adjusted to generate new components, thenew components conform to the actual engineering situation, the repeated family building process is reduced, and the modeling efficiency is greatly improved. In the modeling process, tunnel componentsare classified and coded in a unified mode, uniqueness of a component family is guaranteed, retrieval and updating are convenient, meanwhile, parameters such as sizes involved in the model construction process are designed, consistency and standardization of information expression are achieved by sharing parameter files, and therefore an accurate data basis is provided for tunnel operation and maintenance.

The invention relates to a blanc fixe cementitious rock soil-like material and a method for making simulative rock from the blanc fixe cementitious rock soil-like material and belongs to the technical field of rock material simulation in a geomechanical model test. The blanc fixe cementitious rock soil-like material comprises blanc fixe as a aggregate, bentonite as an auxiliary material, and dilute glue composed of 107 glue and water. The method comprises the following steps of drawing a water-binder ratio-shearing strength f curve by the blanc fixe cementitious rock soil-like material, calculating wet density of different batches of the blanc fixe cementitious rock soil-like materials, and making a small test-piece from different batches of the blanc fixe cementitious rock soil-like materials. The blanc fixe cementitious rock soil-like material has a low cost and no toxicity, can be processed into small blocks easily, and can satisfy requirements of a geomechanical model test on the blanc fixe cementitious rock soil-like material. Simulative rock made from the blanc fixe cementitious rock soil-like material does not influence a model building process and can be made fast so that a model test rate is greatly improved. The blanc fixe cementitious rock soil-like material is suitable for the field of a test on a model built by laying small blocks.

A computer-based process retrieves information organized in documents containing text and / or coded representations of text. The process involves obtaining and labeling a selected set of documents based on content quality, and extracting and representing features from each document in the selected set. The extracted and selected features are modified, and models are constructed using parametric learning algorithms. The constructed models are capable of assigning a label to each document. The model parameters are instantiated using a first subset of the selected set of documents. Parameters are chosen by validating the corresponding model against at least a second subset of the full document set. The constructed models also are capable of assigning labels to similar documents outside a selected subset not previously given to the process of model construction.

A computer-based process retrieves information organized in documents containing text and / or coded representations of text. The process involves obtaining and labeling a selected set of documents, and extracting and selecting features from each document in the selected set. The extracted and selected features are represented, and models are constructed using parametric learning algorithms. The constructed models are capable of assigning a label to each document. The model parameters being instantiated use a first subset of the selected set of documents. Parameters are chosen by validating the corresponding model against at least a second subset of the full document set. The constructed models also are capable of assigning labels and ranks to similar documents outside a selected subset not previously given to the process of model construction.

The invention provides a method and device for analyzing the query requirement. The method comprises the steps that the model building process is carried out, wherein pattern matching trees corresponding to different types are built in advance according to various types of pattern dictionaries composed of query patterns of various types, and query, of various types, obtained according to the pattern matching trees serves as a sample to train a machine learning recognition model; the requirement recognition process is carried out, wherein query input by a user is classified according to the pattern matching trees and the machine learning recognition model. According to the method and device for analyzing the query requirement, the query can be classified by combining the pattern matching trees and the machine learning recognition model, the pattern matching trees and the machine learning recognition model can carry out mutual feedback, and can be continuously improved, and classification accuracy of the query is improved.

A series of algorithms can be applied to an automated machine learning model building process in order to reduce complexity and improve model performance. In addition, the settings and parameters for implementing the automated machine learning model building process can be tuned to improve performance of future models. The model building process can also be monitored to ensure that the current build is based on new information compared to previously builds.

The invention discloses a multi-angle human face detecting method based on weighting of deformable components. The method includes a model building process and a human face detecting process. Different weights are set on different components of a human face according to different contributions of the different components to a detection effect, and effects on important components can be sufficiently utilized assuredly. The problem in terms of large-angle human face detection is resolved by a multi-model fusion method, omission ratio is decreased by means of reducing a threshold value in a detection process, and false detecting ratio is decreased by means of a skin color confirming mode in a model result fusion process. The multi-angle human face detecting method effectively resolves problems of omission and false detection in the human face detecting process based on the components, and can be widely applied to the fields of safety detection, identity authentication, intelligent transportation and the like.

The invention discloses an entity-identification cooperative learning algorithm based on a document type, wherein the algorithm comprises a model building module and a model application. The model building module comprises text-type-based entity-identification classifier building and text classifier building. The model application comprises text type identification and text-type-based entity identification. In the algorithm, an integrated learning algorithm and a cooperative training algorithm are combined; meanwhile, in a data set pre-processing process, text-type-based data set segmentation is carried out, and the variety of data sets is considered. In a model building process, sparse labeled data is utilized as training data, and a plurality of basic algorithms are applied and are integrated by virtue of an integrated learning mode, so that the variety of the algorithms is considered. By combining multiple technologies, and considering the variety of algorithms and data sets, the entity-identification cooperative learning algorithm is guaranteed to obtain satisfying effect on an entity identification task.

The invention discloses a method for improving video resolution through a convolutional neural network. According to the method, a video super-resolution reconstruction model based on the convolutional neural network is constructed according to a video image sequence, and in the model construction process, super parameters, including convolution kernel size, the number of neural network layers, etc. of the convolutional neural network are set based on the characteristics of the video image sequence; and then a single-image super-resolution method is utilized to generate image training weightsused for initializing weight parameters of the video model, redundant information among video frames is fully utilized, multiple frame video images are used as input of the convolutional neural network model, and finally a high-resolution video is obtained through an incremental iteration method. Through the method, the video super-resolution model is high in training speed and high in predictionprecision; and a comparison test verification result indicates that compared with other super-resolution algorithms, the peak signal-to-noise ratio and structural similarity of the images reconstructed through the method have comprehensive optimal results.

The invented method includes the following steps: firstly, defining object problem influence factor, selecting model structure, adopting orthogonal testing design or uniform testing design method to make optimization design for influence factor, sampling range and sampling frequency, then making finite element analysis of designed value to form main sample self-group of the intelligent model, making the main sample self-group undergo the process of characterization treatment, then utilizing the value obtained after characterization treatment to construct intelligent analysis model by utilizing neural network algorithm.

The invention provides a whole vehicle controller environmental model generation method. The method comprises steps: according to a whole vehicle controlled object, a controlled object model is configured; a virtual controller is configured, and input and output signals are selected; a test use case is set, and a driver operation signal is simulated; signals received and sent by the whole vehicle controller are processed; and a generation address is selected, and the environmental model is generated. According to the method of the invention, the whole vehicle environmental model can be generated automatically, a model building process is greatly simplified, model building difficulties are reduced, model accuracy is improved, a tester can be freed from high-load work of simulation environment model building, the tester can really focus on the model simulation test work, and the whole test cycle is shortened.

The invention discloses a fast measuring method of tobacco shred mixing ratio, which takes a light absorbency value corresponding to a 5500 to 4200cm near-infrared spectrum of a diffuse reflection near-infrared spectrogram of finished tobacco shred samples to be measured, a diffuse reflection near-infrared spectrogram of leaf-silk samples to be measured, a diffuse reflection near-infrared spectrogram of cut stem samples to be measured and a diffuse reflection near-infrared spectrogram of slice silk samples to be measured as an independent variable matrix; the maxing ratio of leaf silks, cut stems and slice silks used for producing the finished tobacco shreds forms a dependent variable matrix; and a multivariate curve resolution-alternating least square method is used for calculating the matrix in the third step, thus obtaining the actual mixing ratio of the leaf silks, the cut stems and the slice silks in the tobacco shred samples to be measured. The method overcomes the defects of relatively complicated model building process, relatively small model application range, large work load and the like existing in the prior art, is a method that uses less measuring instruments, has simple process and can realize fast onsite measurement, and can be widely applied to the tobacco industry.

A multi-core shield power cable frequency change multi-conductor transmission line model building method comprises the following steps: measuring a common mode impedance and a differential mode impedance of a multi-core shield power cable; building an equivalent circuit model of the common mode impedance and differential mode impedance; indirectly deriving multi-core shield power cable multi-conductor transmission line frequency change unit length electric parameters; respectively reversing resistance, inductance, capacitance and conductance rational function approximation by polynomial in a pole-residual mode, wherein the system in a stable system; respectively processing when the stable system is connected with a passive network and an active network, thus realizing passivation of elements in a unit length frequency change electric parameter wideband equivalent circuit. The method considers structural asymmetry influences on electric parameter frequency change features in the multi-conductor transmission line model, can effectively isolate surrounding electromagnetic interferences from affecting the impedance measuring result, thus improving the measuring precision; the method is suitable for the frequency change multi-conductor transmission line model building process of the power cable, a communication cable and a split cable models with a simple structure.

The invention discloses an unsupervised entity relationship extraction method based on zero-shot, and belongs to the field of computers. An entity relationship category is judged by extracting triplefeatures in text data and entity relationship type features in a domain knowledge graph and calculating the similarity between the triple features and the entity relationship type features, so that the dependence of a traditional entity relationship extraction method on manual annotation is reduced, and the accuracy of entity relationship extraction is improved. The method comprises the steps of data preprocessing, feature extraction, relation extraction network model training and entity relation classifier training. A convolutional neural network model which is good at capturing sentence information is adopted to respectively extract triple and relationship type features, and finally softmax is used to predict entity relationship type tags. In a model construction process, a sparse taggedcorpus can be used as a training set, and in a test process, the same parameters as those in the training process can also be used for predicting the type of an untagged triad.

The invention discloses a pedestrian detection method based on a weighted part model and selectiv search segmentation. The pedestrian detection method based on weighted part model consists of a weighted part model training part and a pedestrian human face detection part. A model construction process comprises steps of setting various weights on various parts on the basis of the construction process of a transformable part model and according to the different conditions of shielding parts of the pedestrian in a crowded scene, and utilizing a hidden support vector machine to perform training weight. The pedestrian detection process comprises steps of extracting HOG characteristic pyramid from an image, performing selective search and segmentation on the image in order to achieve the fact of employing less effective windows to contain more objects, and, on the basis of that, and adopting cascading detection based on threshold cutting to detect the pedestrian. The invention solves the problem of missing detection in a crowding scnene from two aspects, and reduces the invention and improves the pedestrian detection accuracy.

The invention provides a construction method for a network connection adaptive deep convolutional model. The method comprises the steps that first, weight vectors in a convolutional neural network areorthogonalized; second, connection between layers in the convolutional neural network is deleted according to a norm; and third, an activation function is constructed. According to the method, data relevancy is lowered through weight vector orthogonalization; the connection between layers in the convolutional neural network is deleted based on the p norm, the problem of overfitting in the construction process of the deep convolutional model is effectively lowered, and network connection adaptation is realized; and the activation function is constructed to generalize data information as much as possible on the premise of not changing a parameter scale, and therefore the accuracy of the deep convolutional model is improved.

The invention provides a non-linear process industry fault prediction and inference method. The method comprises: a data pre-processing process, an extended finite state machine (EFSM) model building process, a variable prediction process based on a feedback differential evaluation optimized extreme learning machine (FDE-ELM), and a fault inference process based on the EFSM. The method is advantaged by high model building stability, high fault prediction precision, low algorithm complexity, automatic fault inferring, and visual inference process. The method provides help for ensuring security of process industry, ensuring property and personnel safety, and saving hardware cost.

The invention discloses a construction engineering project progress prediction system and method based on deep learning, which belong to the field of the construction industry. A model building process comprises the steps of extracting historical data; performing data cleaning, correction, supplement and regularization processing on the historical data to obtain to-be-trained data; constructing aninitial prediction model and carrying out training; verifying the model according to the special data, and outputting the initial prediction model passing the verification as a progress prediction model; the progress prediction process comprises the steps of obtaining current actual project information and inputting the current actual project information into the progress prediction model, and predicting predicted project progress information associated with the actual project information. The method has the beneficial effects that a set of progress prediction method suitable for different subdivision fields and scenes is established, historical data are preprocessed and processed through a deep learning algorithm, meanwhile, a special scene is simulated emphatically, a corresponding prediction result is output as the special scene needing to be considered in progress prediction, and an early warning prompt can be given.

The invention provides a three-directional geology modeling method based on sparse borehole points, and belongs to a three-dimensional geology modeling method for coal mine geology. The modeling method comprises steps of 1: constructing borehole database, extracting responding base information, generating comparison diagram of a complete drilling layer and an incomplete layer and using the interpolation algorithm of fixed points; 2: setting mark points of transverse virtual drilling holes and forming the whole drilling hole by using the coordinates as input parameters of interpolation based on a prospecting line profile map according to exploration data of stratum trend; 3: using crosscheck as a statistical means to test precision of models according constructed models; and 4: referring to the contour map of one stratum, comparing corresponding stratum data of models and summarizing similarity until the models are modified to a proper confidence level. In the general background of sparse borehole points of a model construction process, other non-borehole data are added, missed borehole data are virtualized, and according to possible errors of constructed models, judging abruptly changed construction methods, thereby improving construction quality of the models.

The invention is applicable to the technical field of software tests, and provides a model construction method and device for a terminal application. The model construction method comprises the steps of obtaining status sequences from application starting to application closing of the terminal application when the terminal application is running normally, wherein the state in the status sequences comprises an operation order input by a user and an interface pack name which is skipped in the operating process; forming a behavioral training set through the obtained multiple state sequences, wherein any state of the terminal application is included in at least one state sequence of the multiple state sequences; constructing a Markov test model related to the operation of the application based on the behavioral training set. The construction procedure of the model is simple, by adding the state sequence into the behavioral training set, an original test model can be improved, thus the situation that if the functions of the application are modified, the model need to be reconstructed completely is avoided, and the model construction efficiency of the terminal application is improved.

A computer program product stored including machine executable instructions stored on machine readable media, the instructions configured for performing automatic logic build processes and implementing autonomic quality checking, by implementing a method including: providing a model repository for holding at least one component; updating the model repository with at least one component; creating a tag for each sub-component of a selected component of the model repository; associating each tag with a latest version of each respective sub-component; and issuing a component submit notice to identify at least one of a dependency and a priority between selected components. A system is also provided.

The invention provides a CNN based rain measurement identification of a marine radar. The method comprises a model building process, a model optimization process and a model training process. The method is based on a classical LeNet-5 CNN model. A multi-level residual CNN model is established, a training set is established by using samples of light rain, medium rain and heavy rain, and input intothe multi-level residual CNN to train the network model, the loss rate is calculated by using a cross entropy loss function, the minimum loss is obtained rapidly, and achieving a very good convergencestable state by using a batch training mode.

The invention relates to an energy system optimization method based on a comprehensive hub integrated energy model. The method comprises the steps that firstly, comprehensive hubs and connecting relations between the comprehensive hubs and a transmission system are acquired, the integrated energy model is built, then an object function is set according to actual needs, an energy system is optimized according to the object function and the integrated energy model, and optimal input and output of all energy carrier subsystems are obtained under the condition of meeting the object function. Compared with the prior art, the comprehensive hubs are used as basis units for building the model by the built integrated energy model, the model is modularized obviously, and the model building process is simple and flexible.

A simulating method of the object damage process by the impulse force decomposes the execution of the complex object into the summation of the directing damage and the limiting damage of the every component. The process: (1) it builds the geometrical model for the complex object and decomposes it to many components according to the command of the structural mechanics and builds the limiting condition for the components to save the structure character of the object; (2) achieve the directing damage of the stress component of the complex object and define the cracking model then attach the crack in the surface of the component, last to generate the crack side and the shiver according to the stress and crack direction; (3) it finds the limiting components of the present component according to the definition restrain condition in the model building process, then it simulate the damage by the different damage arithmetic according to the difference of the component type. The invention can not only fit the command of the real-time character but achieve fidelity.

The foundation pit excavation and anti-floating anchor rod construction method based on BIM comprises the following steps: 1, an excavation model is built through BIM software; 2, in the excavation model building process, the BIM software is used for extracting the volumes of excavation models before and after hollow shearing, and the difference value of the volumes of the excavation models beforeand after hollow shearing is the excavated earthwork amount; 3, elevation data of the anti-floating anchor rod is summarized; a pre-designed anti-floating anchor rod plane positioning graph is introduced into the BIM software, the anti-floating anchor rod plane positioning graph correspondingly overlaps with the excavation model, the position of the anti-floating anchor rod on the excavation model is determined, elevation marking on the anti-floating anchor rod is performed by the BIM software, and the elevation marking data is summarized and integrated into an EXCLE table; 4, after elevationmarking of the anti-floating anchor rod is completed, a CAD two-dimensional graph and a three-dimensional visualization model graph of the foundation pit model are generated in a segmented mode so asto guide construction; and 5, on-site construction is carried out. The construction quality and the construction period are greatly improved, the rough management mode of the site is changed, and standardized and refined construction is achieved.

The invention relates to the field of computer aided design and discloses a knowledge-based part parameterized modeling method for an automobile welding jig. The method comprises the following steps of extracting design knowledge for the welding jig; performing expression and classification on the design knowledge; by virtue of a software development platform, creating a part model program; adding the design knowledge to the part model program; then designing an interactive interface of the part model program; packaging the part model program; and finally building a part geometric model by using the part model program. According to the knowledge-based part parameterized modeling method for the automobile welding jig, the problems that model building is complicated and parameterized design and design knowledge cannot be effectively organized and expressed in an existing jig part model building process are solved; automatic modeling of jig parts is realized; the model building is quick and accurate; the design efficiency of the welding jig part model is improved; and the product research and development cycle is shortened.

The invention belongs to the technical field of subway shields, and particularly relates to a rapid modeling risk pre-assessment method for a shield to pass through a dense building group. According to the evaluation method, based on limited element analysis software ABAQUS and three-dimensional modeling software GIS, the influence of tunnel excavation of an earth pressure balance shield tunneling machine on a surface building is simulated, and a shield tunneling construction scheme and a building reinforcement scheme are formulated by judging a risk level through a preset evaluation standard. According to the invention, the metrology and the three-dimensional modeling technology are combined, so that in the model building process, the earth surface structure meeting the precision requirement is rapidly built, and the time spent on safety calculation of a traditional building is greatly shortened. According to the invention, detailed risk level judgment standards are provided, preliminary construction scheme optimization ideas such as shield parameter adjustment or foundation reinforcement are provided for designers for different risk levels, and the designers are assisted in formulating and optimizing the construction scheme.

The invention provides a three-dimensional roadway model building method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining roadway modeling data, determining a roadway three-dimensional center line which is a geometric center line of a roadway under three-dimensional projection according to the roadway modeling data, and processing the roadway modeling data according to the roadway three-dimensional center line, and generating a three-dimensional roadway model. According to the method, the automation and intelligence degree of three-dimensional roadway model building can be effectively improved, resource consumption in the model building process is reduced, and the model building convenience and effect are improved.

The invention relates to the field of model construction, in particular to a PHM model construction method and system, a storage medium and electronic equipment. The method comprises the following steps: step 1, collecting to-be-modeled data; 2, an intelligent parameter setting preset sub-process is called to process the to-be-modeled data; and step 3, outputting a processing result as a PHM model. The invention provides a mode of processing data through a standardized process to solve the problem of inconvenient modeling operation. In other words, a user can change the parameters according to requirements, the basic algorithm does not need to be rewritten in the model building process each time to obtain the needed output result, only the algorithm parameters need to be adjusted, then the problem that coupling of data and the algorithm is too tight does not exist, and the effect of improving operation convenience is achieved.

The invention provides an unmanned aerial vehicle three freedom degree model building method. An unmanned aerial vehicle three freedom degree model is composed of an unmanned aerial vehicle data module and an aerial vehicle particle motion dynamics module, the aerial vehicle particle motion dynamics module can be used in the model-building processes of all the unmanned aerial vehicles in a repetitive mode, and therefore the workload of unmanned aerial vehicle three freedom degree model building is reduced. Complex functions such as aerodynamic computation, model resolving and calculating and auxiliary physical-quantity calculation are given to a computer through packaging of unmanned aerial vehicle three freedom degree model software. A user simply needs to set model input, and then can obtain the complete physical quantities of aerial vehicle three freedom degree motion, and the unmanned aerial vehicle three freedom degree model can be used simply and conveniently.