106 results about "Model integration" patented technology

A specific clothing image identification and detection method based on machine learning is provided. The invention relates to an image identification and detection method. The purpose of the invention is to solve the problems of the prior art that a manual detection way cannot realize searching of images of specific kinds of clothing because of a large quantity. The method includes the following processes: firstly, identifying image contents, and obtaining identified image contents; constructing an image classification database, processing images in the image classification database through a self-service sampling method, and obtaining processed images; obtaining an optimized convolution neural network; carrying out model integration; obtaining N individual learning devices, and combining the N learning devices through a simple ballot way; and secondly, utilizing a Faster R-CNN method to detect the identified image contents obtained in the first step. The specific clothing image identification and detection method is applied to the clothing image identification and detection field.

The invention discloses a segmentation method of pathological section unconventional cells based on a multi-scale hybrid segmentation model. The method comprises steps that positive and negative samples are respectively scaled to low-resolution, medium-resolution and high-resolution images, the full convolutional network algorithm is utilized for training to acquire a convergent low-resolution segmentation model, a medium-resolution segmentation model and a high-resolution model; a multi-scale hybrid segmentation model is acquired through fusion by a model integration method; after the effective discriminating area of a new pathological section is processed through utilizing the data enhancement method during testing, the processed effective discriminating area is inputted to the multi-scale hybrid segmentation model, the probability of each pixel in the effective segmentation area is outputted, pixels with probability values greater than the threshold t are taken as abnormal cell pixels and are recorded as 1, remaining pixels are taken as normal cell pixels and are recorded as 0, binary images predicted by the multi-scale hybrid segmentation model are acquired, and post-processingon the binary images is carried out to acquire the final segmentation result. The method is advantaged in that high precision is realized, and a Dice value is above 0.869.

The invention discloses a multi-model meta-synthesis flood forecasting system which comprises a data integration module, component integration module, model integration module, scheme integration module and a result publishing module, wherein the data integration module, the component integration module, the model integration module, the scheme integration module and the result publishing module are sequentially connected. The invention also discloses a method for flood forecast by utilizing the multi-model meta-synthesis flood forecasting system, comprising the following steps of unifiedly treating historical or real-time hydrological data through the data integration to enable the hydrological data to accord with the use specifications and the requirements of the component integration; packaging each set of obtained data into a plurality of components according to the needs of a constructed framework; obtaining a plurality of flood forecasting models by respectively setting up the plurality of components on the basis of the structure requirements of the models and obtaining corresponding flood forecasting results; obtaining a final forecasting result and a final forecasting scheme by treating the plurality of flood forecasting results in a unified way; and publishing the final result. The method realizes rapid setting up of various flood forecasting models and can provide a plurality of schemes and scheme optimizations.

The invention provides a method for evaluating operating reliability of a multi-model integrated aero-engine under multiple failure modes. The method comprises the steps of analyzing the multiple failure modes, establishing a multi-failure mode analyzing model, establishing an alternative evaluation model, judging advantages and disadvantages of each model, establishing a multi-model integrated evaluation model, analyzing the influence from the multiple failure modes to the operating reliability of the aero-engine by applying a Bayesian model averaging method, and carrying out simulation calculation on the operating reliability of the aero-engine by adopting a MCMC (Markov Chain Monte Carlo) algorithm. According to the method provided by the invention, typical failure modes, performance degradation failure, structural strength failure and outburst failure are respectively analyzed, one optimal model is selected from multiple alternative models, multiple models are integrated in one framework by applying a multi-model integration technology for evaluating the operating reliability of the aero-engine, and the evaluation accuracy of the operating reliability of the aero-engine can be increased.

The invention discloses a video rain removing and snow removing method based on multi-scale convolution sparse coding. Under the assumption of a low-rank background, the rain and snow components and the moving prospect in the video are estimated at the same time. Firstly, video data containing rain and snow noise is acquired, and a model is initialized; a generation model of the rain and snow graph is built according to the characteristics of the rain and snow and the video prospect; according to the structural characteristic that the rain and snow are imaged in the video, the moving rain andsnow are repeatable and multi-scale rain strip local blocks on the image, a multi-scale convolution sparse coding model is established related to the rain and snow; a moving object detection model isestablished according to the characteristics of the video foreground sparsity; the model is integrated into rain and snow under the maximum likelihood estimation framework model; a rain-snow video anda rain-removing snow model are applied, so that rain and snow videos and other statistical variables are obtained, and rain and snow videos are output. The invention aims to establish a high-qualityvideo rain-removing snow model based on the rain and snow generation principle and the rain and snow noise structure characteristics, so that the snow removing and snow removing technology can be widely applied in more complicated practical scenes.

A method of generating short-, medium-range and seasonal-timescale weather or climate forecasts by running an ensemble of computer models on a distributed computing system or network. Individual model integrations are interrogated to select those that most closely ressemble observed conditions in the present and recent past and the forecast based on a weighted average of future predictions based on this subset of the ensemble. The selection criteria determining which models are deemed to fit the observations most closely may be adjusted to optimize the use of observations in forecasting specific climate variables or geographic regions in order to develop forcasts tailored to particular applications.

A gas sensor array concentration detection method based on fuzzy division and model integration belongs to the field of a gas sensor array signal processing technology. According to the method, time division of baseline drift data is carried out by a fuzzy clustering method, and a source dataset is divided into multiple sub-datasets with different drift degrees; then, regression models of different training datasets are established to obtain several sub-regression models; an optimal weight set of each sub-regression model is obtained within the training dataset, and center of clustering and optimal weight undergo fitting to obtain an optimal weight fitting function; and during the test phase, fitting weight is calculated on the basis of the optimal weight fitting function and clustering center time, and the sub-regression models integrate forecasted results of data to be tested so as to obtain the final gas concentration value. By the method, a pattern recognition model can be changed adaptively, changes of drifting can be traced, the influence of drifting on concentration detection performance is effectively reduced, and long-term accuracy of concentration measurement is guaranteed.

The UML model integration and refactoring method steps include extending the metamodels, composing an integrated UML model derived from the extended metamodels, defining refactoring opportunities and transformation operations during the integrated UML composition, and applying a set of composite refactorings to remove an identified smell from the integrated UML model.

The invention discloses a design method for an intelligent transformer station model based on CAD (Computer-Aided Design) graph and model integration. According to the method, the design of a secondary loop and the generation of a model file of the intelligent transformer station are synchronously completed; a data object corresponds to the back of each entity, and the model can be operated by a graph, therefore, the consistency of the design of the secondary loop and the model file is ensured, and the problem that the traditional intelligent transformer station is not deep enough is solved; and synchronously the design of the transformer station model can be finished by designers not familiar with 61850 design details, the role and the function of a design organization are strengthened in the establishing of the intelligent transformer station, and the visibility in operation and maintenance of the transformer station is increased; furthermore, the method is combined with the habits of traditional designers and is used for performing graphical loop design based on an AutoCAD platform so that the operation of the designers is convenient; an IEC61970CIM model is used in a primary system to describe the model, therefore, the data supporting other applications can be exported, and the openness of the method is increased.

The invention discloses a municipal administration road BIM design model and a GIS data integration method, and relates to the field of three-dimensional road model integration. The method includes following steps: firstly, acquiring a road BIM design model in a LandXML format, decomposing road model entity elements, and defining a road extended model through the element mapping and an ADE extended frame; executing semantic analysis to discrete sampling data, and converting to acquire the whole road design elements and an object geometric model; finally, according to the different levels of the visual contents, executing the integration exhibition to the road extended model data in three-dimensional software. The method is capable of using the CityGML standard as a foundation, combining the content of the municipal administration road BIM design element, extending the model and building a new road ADE model. Integrity and structural accuracy of macroscopic GIS data and microcosmic road BIM data element integration are improved, and a frame and a feasible path are provided for the integration of the road BIM and GIS data.

The invention provides a municipal engineering construction monitoring method based on a building information model (BIM). According to the method, first, a monitoring point BIM is constructed, the monitoring point BIM, a civil engineering BIM, a municipal pipeline BIM and a surrounding building three-dimensional model are integrated into a monitoring visual model, monitoring data of monitoring points are matched with points in the monitoring visual model and compared with a set threshold value, and different alarm colors are displayed on the corresponding points in the monitoring visual model according to the difference of comparison results; when the monitoring data exceed the threshold value, alarm information of the monitoring points is sent to a related mailbox or mobile equipment. The municipal engineering construction monitoring method based on the BIM is suitable for monitoring all kinds of municipal engineering construction and can visually display planar and spatial position relations between the monitoring point and all kinds of structures, and the information of the monitoring points where potential hazards occur can be sent to the related mailbox and mobile equipment.

The invention provides a fruit and vegetable images classifying and identifying method and system based on model integration. The method comprises the steps that the fruit and vegetable images to be identified are respectively input in each nerve network model of a plurality of nerve network models after training, so that each nerve network model outputs the identification result of the fruit andvegetable images to be identified; integrating the identification result output by each nerve network model, and obtaining the final identification result of the fruit and vegetable images to be identified. The invention is advantageous in that the classifying features of a plurality of network models can be fully developed, and the same sample can undergo multiple classifications of different network models, and finally a classifying identification result after optimization grouping can be generated, which is always more reliable and more accurate than the classifying identification result generated by a single network model.

The invention discloses a computer-aided pulmonary nodule classification method based on migratable multi-model integration to solve the technical problem of low accuracy of an existing pulmonary nodule classification method. The technical scheme includes the steps of training three pre-trained deep convolutional neutral networks (Pre-trained DCNN), describing the textural heterogeneity, shape heterogeneity and global characteristics of pulmonary nodules respectively, performing weighted average for results of the deep convolutional neutral networks, and subjecting the weight of each network to adaptive learning through an error back propagation mechanism so as to improve the accuracy of the pulmonary nodule classification method, wherein, the Pre-trained DCNN migrates the good image characterization capability of the deep convolutional neutral network on a large data set onto a pulmonary nodule classification task so that the training performance of the deep convolutional neutral networks on small pulmonary nodule data is effectively improved. The invention overcomes the technical problem of low accuracy of a classification method based on single type of information, and the accuracy rate can be up to 93%.

The invention discloses a brand new inquiry-related multi-ranking-model integration method. The brand new inquiry-related multi-ranking-model integration method comprises the following steps of: establishing sub-ranking models for each inquiry and inquiry-related document; performing vectoring representation on the sub-ranking models so as to convert a plurality of inquiry-related ranking models into characteristic data and integrate the plurality of ranking models; establishing a new loss function as an optimization target at an inquiry level and a sample level by using a ranking support vector machine as the sub-ranking model; adjusting weight among losses generated by different inquiries by using the loss function; and providing a multiple inquiry-related ranking support vector machine fusion algorithm. Compared with the traditional model, the inquiry-related ranking models can achieve better properties when the inquiry-related multi-ranking-model integration algorithm provided by the invention is applied to actual tasks. The multi-model fusion algorithm provided by the invention can be applied to ranking learning, can also be applied to multi-element classification, sequence labeling and the like, and has a wide application prospect in the fields of information retrieval, network search and the like.

The invention discloses a design method of a complex reaction type system controller, belonging to the field of controller design. In the method, a mode that control software and an embedded hardware platform are in parallel is adopted. The method comprises the following steps: dividing the controller into a logic part and a control law part according to the complex reaction type system attributes; respectively establishing a model for the two parts by using the idea of model driving development, and combining the logic part and the control law part of the controller by using the model integration mode of model importing, thereby forming complete complex reaction type system control software irrelevant to the platform; and finally, converting a personal information management (PIM) model into a platform specific model (PSM) by using a specific operating system adapter, and downloading the PSM model to the designed hardware platform, thereby completing the design of the complex reaction type system controller. By using the design method, the design personnel can concentrate on the construction of the complex reaction type system logic model without considering the editing of the controller software, thereby greatly enhancing the development efficiency of the complex reaction type system controller.

The invention relates to a man-machine asynchronous mechanical ventilation detection method based on a circulating neural network. The method comprises the following steps: a, collecting respiratory records during mechanical ventilation, and marking types of man-machine asynchrony by an expert; b, training and testing a multi-channel circulating neural network model on the basis of the marked respiratory records; c, detecting occurrence of various types of man-machine asynchrony by multi-model integration architecture. Characteristics of respiratory recording pressure waveforms and flow waveforms are fully extracted by adopting the multi-channel model structure, and after the characteristics are fused, the circulating neural network model is used for detecting occurrence of man-machine asynchrony. The occurrence of various types of man-machine asynchrony in the respiratory records can be simultaneously detected.

The invention relates to a virtual fitting method that comprises the following steps of: S1, acquiring the color three-dimensional models of a dressed standard human body and a dress trier, and the three-dimensional model of the color three-dimensional model of an undressed standard human body, wherein the dress trier can visually reflect the figure of his/her body; S2, segmenting the color three-dimensional models of the dressed standard human body to obtain the color three-dimensional model of a garment; S3, calculating a transformation function between the three-dimensional model or the color three-dimensional model of the undressed standard human body and the color three-dimensional model of the dress trier; and S4, integrating the color three-dimensional model of the garment into the color three-dimensional model of the dress trier by using the transformation function to obtain a virtual fitting model. The method and system of the invention make a realistic garment model with a final synthetic effect and enable the garment and the human body to be consistent in illumination effect.

A model integration method and apparatus are disclosed. The model integration method comprises: determining at least two to-be-integrated independent models, wherein each independent model has one output value; acquiring event data sets from historical event records; generating integrated model training samples according to the acquired event data sets; and using a supervised learning algorithm totrain the samples to obtain an integrated model.

The invention discloses a highway engineering process reporting system based on BIM + GIS technology, which comprises the following modules: a, a project quality inspection rule base module; B, BIM model integration module; (c) an automatic acquisition module of that original data; D, intelligent data analysis and management module; E, approval module of process application process; F, intelligentgeneration module of quality inspection data; G. Quality problem early warning and traceability module. The system is advantageous in that BIM + GIS technology is taken as that basis, Relying on thenewest computer and communication technology such as big data, cloud computing, wireless communication and internet of things, the real-time collection, collation and intelligent analysis of construction lofting data, construction record data and test data are realized, and the intelligent evaluation of process application quality inspection conclusion and automatic generation and archiving of quality inspection data are realized. It improves the timeliness and accuracy of process application, promotes the standardization and standardization of construction, and saves human and financial resources.

The invention discloses a transformer substation online secondary equipment standing book information acquisition method. According to the invention, content extension is carried out on an IEC61850 model of equipment such as a protection device, a measurement and control device and the like in the substation, and secondary equipment standing book information is written into a CID model and transmitted to a background monitoring system. The background monitoring system is upgraded, and the acquired secondary equipment standing book information is compiled into a monitoring system model through substation control level MMS communication. Through the technical scheme disclosed by the invention, the monitoring system can update transformer substation secondary equipment information in real time, all-station model integration is carried out on acquired single device information so as to generate a whole plant station secondary equipment model, the acquired information is provided for a power grid information system to use so as to realize overhauling assistance, operating assistance and asset life cycle management of a transformer substation, a panoramic model of the transformer substation and the equipment is constructed, and the informatization, automation and intelligentization level of the transformer substation can be improved.

The invention provides a wind-water-fire joint robust unit combination method. Firstly, an uncertain set is established for the uncertainty of wind power output, and the uncertainty of the wind poweroutput is described in three aspects of a wind power output prediction interval, a time smoothing effect and a spatial cluster effect. Then, a mixed integer linear programming model for water and power dispatching is built by using a linearization method, and the model is integrated in a robust unit combination model considering wind and power dispatching to obtain a wind-water-fire joint robust dispatching model. Finally, a robust optimization model in two phases is solved by adopting a C&CG method. Compared with a traditional robust unit combination model, the wind-water-fire joint robust unit combination model has the advantages that the operation cost is reduced, and the starting number and the running time of thermal power units are reduced, so that the carbon emission is reduced, andthe environmental benefit is achieved; the wind power consumption capacity of a system is increased due to the addition of water power; and the problem of a unit combination containing cascade hydropower stations can be solved within a reasonable time range.