355results about How to "Solve the characteristics" patented technology

The invention provides an mVSV virus vector, i.e., attenuated mVSV obtained after multiple modification mutations occur to an M protein amino acid site of a wild Indiana strain VSV, and an optimized heterologous antigen gene is preferentially integrated to a double cloning site area of an mVSV packaging core plasmid pmVSV-Core at the same time. The mVSV virus vector vaccine comprises a heterologous antigen gene which fuses or embeds a target virus between G and L genes of an mVSV vector envelope, wherein the antigen gene comprises an enveloped and embedded antigen gene encoding the target virus, an embedded combination antigen gene or a fused antigen gene; the mVSV virus vector is embedded or fused with a dominant antigen of spike protein S of an SARS-CoV-2 pathogen; the dominant antigen is preferably selected from a receptor binding domain of spike protein S, namely RBD; and a COVID-19 vaccine based on mVSV mediation is formed. The vaccine has good prevention or treatment effect on COVID-19 infected people.

The invention discloses a method for identifying vehicle identity by using vehicle position information, which comprises the following steps: storing registration information and management rule information of legal vehicles in a management server, allocating vehicular positioning equipment with unique identification number for the legal vehicles, and establishing association between the identification number and corresponding vehicles in the management server; and comparing the appearance characteristic information of the audited vehicles and the registration information, comparing the position time information fed back by the audited vehicles and the position time information when acquiring the appearance characteristic information of the vehicles in field, and judging whether the identities of the vehicles are normal according to the comparison results. The method makes full use of the position time information of the vehicles, realizes identification of vehicle identity, solves the problems whether the appearance characteristic information of the vehicles and the registration information are accordant and whether the accordant vehicles are in the audited field, automatically identifies the real identities of the vehicles against rules, identifies the vehicles which do not stop at toll sections, automatically charges and automatically discovers the suspected vehicles against traffic rules such as illegal operation, over speed, running against the rules and the like.

The embodiment of the invention discloses a malicious traffic detection implementation method and device based on deep learning. The method can include: obtaining the flow session of malicious code through dynamic sandbox technology; mapping the traffic session of malicious code to a gene map and extracting the map features; using the map features of traffic session for clustering and marking theclustering results by a malicious code family; training the preset depth learning model by using the tagged gene map of the malicious code family to establish the malicious traffic detection model; using the malicious traffic detection model to detect real-time network traffic, and realizing malicious traffic detection. Through the embodiment, many problems existing in the conventional detection technology such as artificial feature extraction difficulty, privacy disclosure, encryption and confusion difficult to identify, machine learning feature manual extraction and the like are solved to acertain extent, and the method has the characteristics of high robustness, high speed, high accuracy, low false alarm rate, cross-platform detection and the like.

The invention discloses a deep well tunnel cable anchor rigid-flexible coupling support and surrounding rock overall reinforced support method which is characterized by firstly excavating a tunnel according to the shape of cross section of the tunnel, and spraying concrete on the surface of surrounding rocks of the cross section formed by excavation; adopting a steel bar mesh and a high-strength prestressed anchor for supporting the surrounding rocks of the tunnel in time; implementing high-strength prestressed anchor cables at the center and two spandrels at the top part of the tunnel; re-spraying the concrete on the surface of the surrounding rocks of the tunnel; mounting a grouting anchor pipe on a top plate and two side surrounding rocks of the tunnel after the tunnel is excavated for 15-20d, and injecting cement slurry into fracture regions of the surrounding rocks of the tunnel; excavating a baseplate of the tunnel after the tunnel is excavated for 25-30d, mounting the grouting anchor pipe, injecting cement slurry liquid into a rock layer of the baseplate; and finally using the concrete to backfill the baseplate of the tunnel flat. The method can solve the problems of large deformation of the deep well tunnel and serious floor heave, avoid the repair work of the tunnel and have significant economic benefits.

The invention discloses a claw beam lifting mechanism for a forging manipulator, applied to forging industry. The claw beam lifting mechanism mainly comprises a front lifting part, a rear lifting part, a synchronous connector rod, a lifting cylinder, a pitching cylinder and a buffering part, a front fixing shaft in the front lifting part, a rear fixing shaft in the rear lifting part and the lower end of the lifting cylinder are respectively connected with a large machine frame through hinges, the synchronous connecting rod is arranged overhead and always moves in a range above a connecting line of a front fixing hinge point and a rear fixing hinge point of a claw beam lifting and pitching mechanism and the large machine frame, front ends of two buffering cylinders are respectively connected with tow front lifting connecting rods through hinges, the rear ends of the two buffering cylinders are respectively articulated with two ends of a compensating beam, and the middle part of the compensating beam is connected with the middle part of the large machine frame by a hinge through a beam fixing shaft. The claw beam lifting mechanism has novel and scientific design concept, reasonable stress application, high movement precision, simple structure, low manufacturing cost and is convenient to install, adjust and maintain. The invention not only reduces the manufacturing cost of the lifting mechanism under the condition of ensuring the movement precision, but also thoroughly solves the problems that the front fixing shaft and a shaft sleeve matched with the front fixing shaft are worn easily and a side plate of a machine needs to be thickened specifically.

The invention discloses an image detection method and system based on a convolutional neural network, and the method comprises the following steps of collecting and obtaining a preset number of labeled sample images, and carrying out the preprocessing of the labeled sample images, and obtaining a denoised target area image with a label; performing feature extraction on each labeled target area image to obtain a 3D feature matrix of each labeled target area; training the 3D convolutional neural network weak entity detection model by using all the obtained 3D feature matrixes to obtain a trained3D convolutional neural network weak entity detection model; and performing preprocessing and feature extraction on the to-be-detected picture to obtain a 3D feature matrix of the to-be-detected picture, inputting the 3D feature matrix of the to-be-detected picture into the trained 3D convolutional neural network weak entity detection model, and outputting a detection result of the to-be-detectedpicture, so that the weak entity target detection can be completed more accurately and quickly.

The invention provides a method for precisely editing a genome DNA sequence. A TALEN, CRISPR/Cas9 and other gene targeting technologies and homologous recombination technologies are utilized in a combined mode for precisely editing the genome DNA sequence. According to the method, the technical problem that in the prior art, a specific sequence cannot be deleted is solved, the specific base sequence can be seamlessly inserted or deleted in a fixed point mode, and seamless free modification of the genome DNA sequence of eukaryocyte is achieved for the first time.

The invention relates to a direct calibration method for a line structured light vision sensor. The method comprises the following steps: establishing a target formed through standard gauge blocks and by utilizing a three-dimensional precision translation bench; projecting structured light plane to the target through the line structured light vision sensor to form feature points on the target; establishing a space coordinate system on the target; collecting target images and extracting the feature points; establishing a search index table including establishing a direct mapping relation between image coordinates of the feature points and space coordinate system coordinates of the feature points; carrying out calibration on the line structured light vision sensor; and based on the image coordinate of a point to be calibrated, searching the space coordinate system coordinate, which is obtained directly or obtained by utilizing a least squares fitting method, of the point in the search index table, so that the direct calibration of the line structured light vision sensor can be realized. The calibration method is used for the line structured light vision sensor which belongs to the small vision field high precision measurement application field. The calibration method is simple and direct, easy to realize, high in precision and can meet the need of measurement.

The invention discloses a self-coding network and a training method thereof, and a method and a system for detecting abnormal power consumption, wherein the training method comprises the following steps: splicing the sample power data by a sliding window to obtain a training sample set, and marking the training sample set containing the surveyed user to obtain a labeled sample, wherein the unlabeled training sample is a non-labeled sample; carrying out unsupervised training on the self-coding network by unlabeled samples, and obtaining the initialization parameters of the self-coding network,then taking the discrete class tags obtained from the coding layer of the coding network as classifiers, carrying out supervised training on the classifiers by the labeled samples, and updating the parameters of the coding layer to obtain the trained self-coding network; then, utilizing the trained self-coding network to detect the power data of the user to be measured, and judging whether the user to be measured abnormally uses power. The invention can mine abnormal information in low-density electric power data, avoid noise data interference, and improve abnormal detection accuracy.

The invention discloses a family member mix identification method and system based on television watching log mining. The method comprises the steps of acquiring log data of television programs watched by a user, wherein the log data comprises following data content: user operation time, program type, and watching time of the program type; dividing the log data by time period; extracting the program type and the proportion of the watching time of the program type from each period of log data and taking the program type and the proportion of the watching time of the program type as a feature of a family member; clustering according to features of family members; and determining the mix of family members by use of a clustering result. Through adoption of the method, the problem in the prior art that user preferences cannot be analyzed through a single account in the intelligent television field is solved. The invention also discloses a family member mix identification system based on television watching log mining.

The invention provides an indoor positioning method for a mobile robot based on multi-sensor fusion. The milemeter position data based on track plotting is adopted for compensating for an unrecognized similar environment in laser matching positioning. According to the invention, the sensor positioning data based on an inertia unit and a milemeter is fused, and the laser radar matching positioning is also considered as a reference index, so that the accumulative error of the milemeter is reduced, and the defect that a short-distance laser radar cannot be matched and positioned under a single characteristic environment is overcome, so that more accurate positioning data can be supplied for map drawing and navigation of the robot.

The invention belongs to the technical field of radar emitter signal modulation recognition, and particularly relates to a modulation recognition method for extracting time-frequency image features byjoint entropy and pre-training CNN. The method includes the following steps: firstly, performing time-frequency transformation on 9 types of radar signal sets to be identified to obtain a time-frequency image; and then based on a pre-training convolutional neural network model imagenet-vgg-verydeep-19 provided by the MatConvNet official website, constituting an FT-VGGNet-fc6 feature migration extraction module from an input layer to an fc6 full connection layer; and then, sending an adjusted image to the feature migration extraction module, and outputting time-frequency image features of radar signals; performing graying on the adjusted image, and manually extracting the Renyi entropy of the processed image; and then, dividing a training set and a test set according to a certain proportion, and selecting the training set to train an SVM classifier; and finally, adopting the trained SVM classifier to recognize the training set of time-frequency images, and adopting a data set composedof 9 types of radar signals with multiple signal-to-noise ratios to verify a recognition rate of an FT-VGGNET-fc6-SVM classifier.

The invention discloses an expansion full-convolution neural network and a construction method thereof. The neural network comprises a convolution neural network, a feature extraction module, and a feature fusion module connected in order; the construction method comprises the following steps: selecting the convolution neural network; removing a full-connection layer and a classification layer forclassifying in the convolution neural network, and only leaving the middle convolution layer and a pooling layer, and extracting a feature map from the convolution layer and the pooling layer; constructing a feature extraction module, wherein the feature extraction module comprises multiple expansion upper-sampling modules connected in series, each expansion upper-sampling module respectively comprises a feature map merge layer, an expansion convolution layer and a deconvolution layer; and constructing a feature fusion module, wherein the feature fusion module comprises a dense expansion convolution block and a deconvolution layer. The expansion full-convolution neural network disclosed by the invention effectively solves the feature extraction and fusion problem in the convolution neuralnetwork, and can be applied to a pixel-level labelling task of an image.

This invention discloses a method for realizing 3-D video result including: converting a video image to grains and mapping them onto a preset 3-D pattern to realize the 3-D video result. This invention also discloses a device for realizing it.

The invention discloses an asymmetrical bolt-net-cable combined supporting method for a special-shaped recovery tunnel of a large inclination angle coal bed. The method is characterized by comprising the following steps of: digging the tunnel according to the shape of the tunnel section, and supporting tunnel top plates, tunnel high-cut surrounding rock and tunnel low-cut surrounding rock by using prestressed anchor rods, steel bar nets and steel strips; arranging top plate prestressed anchor cables on the tunnel top plates; for the solid coal recovery tunnel, arranging high-cut prestressed anchor cables on the tunnel high-cut surrounding rock when the tunnel is dug for 5 to 7 days, and arranging struts in the center of the tunnel towards one side of the low-cut surrounding rock; and for the small coal post reserved recovery tunnel, arranging low-cut prestressed anchor cables on the tunnel low-cut surrounding rock when the tunnel is dug for 5 to 7 days, and arranging struts in the center of the tunnel towards one side of the low-cut surrounding rock. The combined asymmetrical supporting method has good application effect of stability control of the special-shaped recovery tunnel of the large inclination angle coal bed, the repair frequency of the tunnel is greatly reduced, and the supporting cost of the tunnel is reduced.

The invention relates to a system and a method for achieving multi-priority special vehicle semaphore control based on RFID. The system comprises a plurality of RFID receiving ends, information conversion equipment and traffic semaphore controller, wherein the RFID receiving ends are arranged in positions in distance preset by the system from parking lines in the inlet direction at intersections in corresponding distance; the information conversion equipment is used for converting vehicle information received by an RFID receiving end into main node signals which are transmitted to the traffic semaphore controller; the traffic semaphore controller is used for controlling traffic signals according to the main node signals. By adoption of the system and the method for achieving multi-priority special vehicle semaphore control based on RFID, an RFID technique can be combined with a traffic signal control technique, public traffic vehicles and other special vehicles can be controlled and released according to different priorities, the traffic control effect is remarkable, the times that carrier vehicles stop at the intersections are reduced, remarkable benefits can be made, and the system and the method are applicable to large-scale popularization and application.

On a conventional T-shaped glass chip, a permanent magnet is utilized to fix and decorate a nanometer magnetic bead of a DNA probe in a separation channel of a chip; the rapid, sensitive, high-specific, high-stability, low-cost separation detection on the DNA is realized by utilizing electric drive sampling and combining a laser induced fluorescence detector or an electrochemical detector, thus solving the defects that the time of separating and detecting the DNA by utilizing a microfluidic chip is long, the separation effect is poor, the sensitivity is low, the specific is poor and the stability is poor and the like; the interface manufacturing is avoided by an electric sampling manner; the influence of dead volume is removed; and the sample use amount is reduced in a narrower and uniform micro pipeline, and meanwhile the lower detection limit is ensured.

Embodiments of the present invention provide a method and an apparatus for extracting image features. The method for extracting image features comprises the steps of: acquiring a plurality of cluster centers from an image data set to be classified to serve as low-level feature extractors by using a clustering algorithm; performing convolution for each image in the image data set by using the plurality of low-level feature extractors, and generating a plurality of convolution images for each image, wherein the number of the convolution images is the same as the number of the low-level feature extractors; respectively performing threshold processing for the plurality of convolution images to obtain a plurality of sparse images; performing low-level feature integration for the plurality of sparse images to obtain a plurality of integrated images; and performing middle-level feature extraction for the plurality of integrated images to obtain middle-level features. The method in the embodiments of the present invention can adaptively extract image features and can achieve high extraction efficiency.

The invention provides a device and method for providing end-to-end quality of service (QoS) assurance in service network, and the device is implemented by SRM and mainly comprises: service processing unit, resource management unit and service interface unit. And the method comprises: SRM receives user service request and processes the user service according to the monitored resource utilization condition of application server and QoS requirements of user service, and returns the processing results to the service request sender side. And the invention can assure end-to-end QoS in service network, and raises utilization efficiency of application server in service network.

The invention discloses a method and a system for constructing and inquiring a knowledge graph in the field of science and technology, and belongs to the technical field of science and technology management information. According to the method and system for constructing and inquiring a knowledge graph in the field of science and technology, a user sorts data in a relational database, sorts out object, relation and attribute information to be defined, defines the object, relation and attribute information through data modeling, and defines mapping information from a relational database data table to objects, relations and attributes; and the relational database data into is extracted and converted into objects, relations and attribute instances, and the objects, the relations and the attribute instances are stored in a graph database, and graph query analysis is carried out through attribute value matching. According to the method and system for constructing and inquiring a knowledge graph in the field of science and technology, the value exploration and information discovery requirements of the user on the big data gold mine can be better met, and the popularization and application value is very good.

The invention discloses a hybrid cloud time series data-oriented anomaly detection method based on an ensemble learning technology, and relates to the technical field of data processing and anomaly detection. The method comprises the following steps: obtaining and preprocessing historical data; establishing a time series data anomaly detection model; jointly training the three basic prediction models; prediction the timing data anomalies. According to the method, through analysis and processing of log time series data and an integrated learning framework based on STL, ARIMA and LSTM, a unifiedand integrated anomaly detection model for coping with a hybrid cloud complex scene is established, anomaly detection and prediction of service indexes are carried out in time, possible abnormal changes are found in advance, fault emergency processing is carried out, and the probability of service fault risks is reduced.

The invention discloses a method for processing thin wall welder cases of revolving body structures, belonging to the technical field of aero-engine metal plate forming and welding. According to the invention, a whole pre-deformation welding/forming method is used for solving the problems that sizes of mounting edges are out of tolerance, coaxiality of front and rear mounting edges with main bodies after welding is out of tolerance and residual stress distribution of welding components is unbalanced. The method provided by the invention comprises the following steps of: carrying out technological structure welding deformation allowance compensation design on a mounting edge and a main body of a thin wall welder case before the welding according to the plastic deformation characteristics of materials; plastically forming an expansion block according to an expander work part, size and final structural size requirement of parts; carrying out automatic argon arc welding on the mounting edge and the main body; assembling the welded welder case and expansion block on the expander work part to carry out expansion plastic forming and controlling the expansion forming size through adjusting pressure of the expander; and unloading the pressure of the expander, recovering the expander work part and unloading the parts to complete the welding/forming process.

The invention belongs to the technical field of radar signal analysis, and discloses an LFMCW signal rapid detection and estimation method. Quasi-coherent accumulation is performed on the LFM part of each period of the same modulation frequency in LFMCW signals according to PFRFT energy so that the signals are enabled to have great energy concentration characteristic in the PFRFT domain, i.e. the PFRFT of the signals has an obvious peak value in two sections. Firstly rapid period fractional order Fourier transform is performed on the linear frequency modulation continuous wave signals, the maximum modulus square of period fractional order Fourier transform act as a target function to establish an optimization model, and the optimization model is solved by using a genetic algorithm so that the optimal parameter estimation of the signals can be obtained. Four-dimensional parameter searching is performed on the PFRFT energy of the linear frequency modulation continuous wave signals by using the genetic algorithm so that detection and estimation of radar weak signals can be realized, the method is especially applied to the design of a signal analysis system of low interception probability radar and thus the theoretical and project application of the PFRFT algorithm can be realized.

The invention relates to a wind turbine planetary gear box fault diagnosis method based on ACGAN. The method comprises the following steps: collecting a vibration signal of a planetary gear box as a diagnosis sample; dividing the diagnosis sample into a training set sample and a test set sample according to a set proportion; inputting the training set sample into the ACGAN for adaptive training, obtaining parameters of a discriminator network and a generator network in the ACGAN until the ACGAN reaches Nash equilibrium, and storing the trained parameters of the discriminator network and the generator network in the ACGAN; taking the trained ACGAN as a fault diagnosis model, inputting the ACGAN into a test set sample, generating a realistic sample by a generator network, and adding the realistic sample into a diagnosis sample; and enabling the discriminator network to output a gearbox fault diagnosis result. Compared with the prior art, the method has the advantages that the original data training network can be directly used, the feature vectors are automatically extracted, and the accuracy of model recognition and classification is high; meanwhile, the generalization ability is extremely high, and fault diagnosis can be effectively carried out on the planetary gear box of the wind turbine generator.

The invention discloses an experiment device used for researching effects on fish behaviors imposed by light and the application thereof. The experiment device comprises a water tank, a double-layered sleeve, a transparent waterproof glass reinforced plastic lampshade , an LED lamp pipe set, a motor, a circuit control system and a camera. According to the invention, by arranging a PVC cylinder of the double-layered sleeve, the experiment is carried out when fish freely swims in an experiment water tank, so via light transmitting holes in the PVC cylinder, experimental fish can freely swim in and out of the cylinder, and an ideal habitat is provided for the experimental fish, thereby simulating the free choices to light wave length and illuminance of the fish in natural sea and industrial aquaculture conditions in a more real manner. Thus, data deviation caused by harmful effect to the fish behaviors by improper structure and operation is avoided. Via the experiment device, many kinds of experiments on effect on fish behaviors imposed by light can be well finished, and the experiment device can satisfy many experimental objectives and requirements for researching effect on fish behaviors imposed by light.

The invention discloses a vision-based pose stabilization control method of a moving trolley, which fully considers about a kinematics model and a dynamics model of a trolley and a camera model. The vision-based pose stabilization control method comprises the following steps of: respectively obtaining an initial image and an expected image at a starting pose position and an expected pose position through a camera, and obtaining an existing image in a movement process in real time; by utilizing an antipode geometric relation and a trilinear restrain relation among shot images, designing three independent ordered kinematics controllers based on Epipolar geometry and 1D trifocal tensor by utilizing a three-step conversion control policy; finally designing a dynamic conversion control rule by taking outputs of the kinematics controllers as the inputs of the kinematics controllers by utilizing an retrieval method so that the trolley quickly and stably reaches an expected pose along a shortest path. The invention solves problems in the traditional vision servo method that the dynamics characteristic of the trolley is not considered during pose stability control and slow servo speed is slow, and the vision-based pose stabilization control method is practical and can enable the trolley to quickly and stably reach the expected pose.

The disclosed drug detector based on animal olfaction comprises: a CCD camera, a brain stereotaxic instrument, an electrode, a pre-amplification and filter circuit, a multiarm maze, and computer control and analysis system. Wherein, according to animal brain olfaction cortex neuron electric action, letting animal produce specific memory to volatile drug; random putting drug in small rooms of the maze to give prize stimulation for specific brain area and reinforce the memory till animal action includes statistic meaning, collecting relative induced brain electric information to input computer for recognition and analysis and form training database.

The invention discloses a method for establishing a crystal plastic constitutive model in a face-centered cubic material fatigue process. According to the method, the problem that mesoscopic scale plastic evolution and material microstructure characteristics cannot be considered in the process of calculating material damage and fatigue failure is solved, and a more accurate and innovative mesoscopic crystal modeling method, a crystal plastic constitutive model and related stable and efficient calculation methods are provided. According to the crystal plastic constitutive model and the calculation method, the normal direction and the sliding direction of the crystal sliding face are considered, the back stress effect is considered through the motion and isotropic hardening criterion, and finally the face-centered cubic metal material crystal plastic constitutive model of the crystal sliding and dislocation density and plastic shear strain rate function relation is established. The influence effect and contribution degree of the microscopic characteristics of the material on the damage and fatigue life of the material can be studied from the mesoscopic scale, important theoretical guidance and technical support are provided for studying the damage evolution and fatigue failure process of the material, and the method has important scientific significance and engineering application value.

The invention discloses a dynamic pore-pressure static sounding probe for detecting sludge. A coaxial cable (1) is used for transferring a signal and data; an analogue-digital converter (2), an acceleration transducer (3) and a gradiograph (4) are arranged on an upper half section of the probe; a side-wall friction cylinder (5) is arranged on a lower half section of the probe; a side wall frictional resistance sensor (6), a pore water pressure sensor (7), a disc resistance sensor (8) and a temperature sensor (9) are arranged in the side-wall friction cylinder (5); a disc (10) is connected under the side-wall friction cylinder (5); a filtering net (11) is arranged at the center of the disc (10); the filtering net (11) is also located at the center of the side-wall friction cylinder (5); and a water inlet channel is formed between the filtering net (11) and the pore water pressure sensor (7). The application of the probe has the characteristics of in-situ, directness, quickness, accuracy, economical property, and the like, and a powerful testing tool is supplied to geotechnical engineering practice.

The invention discloses a three-dimensional reconstruction method based on multi-view self-adaptation characteristic registration. The method comprises the following steps that: A, seed points in each two-dimensional picture are selected; B, a region growing method is adopted to perform image segmentation on the two-dimensional picture so as to obtain the boundary point set of each image of each picture; C, through combining the boundary point sets of the segmented images and a polar geometry constraint theory, three-dimensional registration is performed on two images of which the angles are close to each other in a self-adaptation manner; and D, and three-dimensional reconstruction is performed according to the parallax images of boundary points. According to the three-dimensional reconstruction method based on the multi-view self-adaptation characteristic registration of the invention, the region growing method for image segmentation and the polar geometry constraint theory are fully utilized, such that requirements for characteristic point selection and three-dimensional registration can be satisfied. The three-dimensional reconstruction method is applicable to the implementation of three-dimensional reconstruction systems.