155results about How to "Fully capture" patented technology

A gasket system for sealed interfaces utilizes a carrier ring having inner and outer compression stops to account for and control flange rotation effects in order to achieve a uniform and constant gasket seating stress across the sealing region. The inner compression stop provides a fixed fulcrum for flange rotation. The inner and outer compression stops have different heights, dependent upon the characteristic flange rotation, that allow the orientation of the sealing material to be angularly matched to the flange side profile and that allow the sealing material to be fully captured within the sealing region upon full compression, minimizing creep relaxation. A visual compression indicator on the circumference of the carrier ring permits assembly of the interface without the use of specialized or calibrated tools.

The invention discloses a refining flux used for purifying and processing aluminum or aluminum alloy melt, and its preparation method. The invention is characterized in that the raw material formula of the refining flux is a powder mixture composed of the following components by mass: 25 to 35% of NaCl, 25 to 35% of KCl, 15 to 25% of Na3AlF6 and/or K3AlF6, 4 to 10% of Na2SO4 and/or K2SO4, 3 to 10% of CaCO3 and/or Na2CO3, 3 to 6% of REF3 and/or RECl3; wherein, a mass ratio of NaCl to KCl is 1:1; RE is Ce or La. The refining flux can be fully diffused after entering the melt without adsorption agglomeration and pore plugging; and has high efficient capability for producing dry slag without viscous aluminum melt; even the refining flux has a slight reaction with aluminum melt, the reaction products can be removed with slag; the refining flux provided in the invention has high efficiency of gas and slag removing.

An image capturing system utilizes a plurality of image capturing apparatus to capture first image data for an object. A processor is utilized to combine the first images to produce first three-dimensional digital image data. The processor then utilizes a second algorithm to combine the intermediate digital image information to produce a complete three-dimensional digital image of the object. The three-dimensional image may be viewed on a display and viewed from any point on a three dimensional surface surrounding the object. The object can be non-stationary.

The invention discloses a session recommendation method based on a space-time sequence diagram convolutional network. The method comprises the following steps: S1, modeling all session sequences intoa directed session graph; S2, constructing a global graph by taking common commodities in the session as links; S3, embedding an ARMA filter into a gated graph neural network, extracting a topologicalgraph signal which changes over time from the graph model, and obtaining a feature vector of each node involved in the session graph; S4, obtaining global preference information from historical sessions of the user by adopting an attention mechanism; S5, obtaining local preference information of the user from the last session clicked by the user, and obtaining final preference information of theuser in combination with the global preference information; S6, predicting the probability of possible occurrence of the next clicked commodity in each session, and giving a Top-K recommended commodity. According to the method, rich context relationships of clicked commodities can be captured from the global graph, global and local preferences of the user are accurately learned, the time attenuation effect of historical preferences of the user on current preferences is effectively evaluated, and accurate commodity prediction is provided.

The invention discloses an electroluminescence device, which structurally and sequentially includes a substrate/a conducting layer/a hole injection layer/a hole transmission layer/an electronic blocking layer/a luminous layer/a hole blocking layer/an electronic transmission layer/an electronic injection layer/a cathode layer, wherein the luminous layer adopts a composite layer structure; the composite layer structure sequentially includes a first blue light luminous layer/a first quantum well/a red light-green light phosphorescence luminous layer/a second quantum well/a second blue light luminous layer; the first blue light luminous layer is prepared on the surface of the electronic blocking layer; and the hole blocking layer is prepared on the surface of the second blue light luminous layer. According to the electroluminescence device provided by the invention, the first and the second blue light luminous layers are placed on the two sides of the red light-green light phosphorescence luminous layer respectively, so that the illumination of the red light and the green light can be completely excited by the emitted blue light, the energy utilization ratio is improved, and further the luminous efficiency can be enhanced.

The invention discloses an asymmetric multi-modal fusion saliency detection method based on an attention mechanism. The method comprises: inputting an RGB image and a depth image of an original three-dimensional image into a convolutional neural network for training to obtain a corresponding saliency detection image; calculating a loss function between a set formed by the saliency detection imagegenerated by the model and a set formed by corresponding real human eye gaze images to obtain an optimal weight vector and an offset term of the convolutional neural network classification training model; and inputting the three-dimensional image in the selected data set into the trained convolutional neural network model to obtain a saliency detection image. According to the saliency detection method, RGB and depth map features are fully extracted by adopting an asymmetric coding structure, rich image information of RGB is effectively utilized after an internal sensing module is added, channel and spatial attention mechanisms are added, the expression of a saliency region and saliency features is enhanced, and the detection accuracy of visual saliency detection is improved.

The invention discloses a swing wing type ocean current energy power generating device. The swing wing type ocean current energy power generating device comprises a supporting frame, a motion mechanism is connected to the supporting frame and connected with an energy extracting mechanism, the energy extracting mechanism comprises a crank and rocker mechanism which transmits energy extracted from the motion mechanism to an energy conversion mechanism, and the energy conversion mechanism comprises a power generator; and the motion mechanism is connected with an attack angle adjusting mechanism, and the attack angle adjusting mechanism adjusts and controls an attack angle degree of the motion mechanism so as to enable the motion mechanism to achieve preset pitching motion. The hydrofoil attack angle adjusting mechanism is arranged, so that the problem in control over the hydrofoil attack angle degree is solved; and as the hydrofoil pitching motion is accurately controlled by monitoring the position of a swing rod, the swing wing type ocean current energy power generating device can adapt to different water flow speeds, kinetic energy of the water flow is sufficiently captured, and the power generation efficiency and power are improved. A transmission system composed of a four-link mechanism and a synchronization belt in a transmission manner is used, so that the structure is simple, mechanical losses are small, and the conversion efficiency is high.

The invention discloses an attack perception method, device and equipment based on honeypot induction and a medium. The method comprises the following steps: filtering a captured data package to obtain a first target data package, and outputting the first target data package to a user mode process; determining a target communication induction strategy corresponding to the first target data packetin a user mode process, and communicating with a source Internet protocol address corresponding to the first target data packet based on the target communication induction strategy to capture at leastone second target data packet sent by the source Internet protocol address; and determining the first target data packet and/or the second target data packet corresponding to the first target data packet as a to-be-analyzed data packet, analyzing the to-be-analyzed data packet based on a preset analysis rule set, and analyzing an attack behavior according to an analysis result. The system has thedual advantages of an intrusion detection system and a honeypot system, attack information can be captured more comprehensively, and the safety of the network is protected.

The invention discloses an osteochondral stent as well as a preparation method and application thereof. The osteochondral stent comprises a cartilage layer, an adhesion layer and a bone layer, wherein two sides of the adhesion layer are respectively connected with the cartilage layer and the bone layer; one or more of the cartilage layer, the adhesion layer and the bone layer are of a porous structure. The osteochondral stent can be used for repairing osteochondral defects, can fully capture cells, has a strong osteochondral injury repairing function, and is simple and convenient to operate.

The invention discloses a microblog friend recommendation method based on a heterogeneous information network. The microblog friend recommendation method comprises the following steps: S1, constructing heterogeneous information network data of a microblog user; S2, extracting meta-paths in the heterogeneous information network, and converting the heterogeneous information network into a homogeneous information network; S3, obtaining vector representation of the user by using a graph convolution network; and S4, calculating the similarity between different users and carrying out similarity fusion. By using the heterogeneous information network and the graph convolution network, rich user behavior information is captured, and high-performance user vector representation is obtained, so that the recommendation accuracy is effectively improved.

The invention provides a cellular network traffic prediction method based on a three-dimensional convolutional neural network. The cellular network traffic prediction method comprises the following steps: modeling network traffic data into a three-dimensional tensor input form to obtain a three-dimensional network traffic data model; obtaining training set data and test set data according to the three-dimensional network flow data; constructing a basic three-dimensional convolutional neural network; performing three-dimensional convolutional neural network training on the short-time dependence data to obtain short-time features, and performing three-dimensional convolutional neural network training on the long-time dependence data to obtain long-time features; carrying out fusion training on the short-time features and the long-time features to obtain a feature matrix, and taking the feature matrix as the output of the basic three-dimensional convolutional neural network to form a training model; and predicting the network traffic data to be predicted by using the training model to obtain a network traffic prediction result. According to the prediction method provided by the invention, the short-term correlation and the long-term tendency of the network traffic data are considered at the same time, and the characteristic correlation of the network traffic data in time sequence is captured.

The invention discloses a fault recording device, a decentralized fault recording system and a method. The scheme is a recording device which is freely arranged according to the input channel of analog quantity and switch quantity, dual-network redundant connection is conducted between devices, network synchronization can be carried out through measurable time delay, inter-device network triggering is performed, each device records and sends its own sample data, and also records sample data sent by other devices, and access from any device can view the technical realization scheme of the decentralization of all data. As the invention is different from the traditional fault recording realization scheme, the invention has the following characteristics: configuring by channel number saves screen cabinets, double network redundancy improves the reliability of the connection, the delay measurable network timing saves the connection lines and interfaces, the network synchronous triggering realizes the panoramic capture of disturbing data, the redundant records of each device improve the reliability of the data, and the access to all data at any point solves the problem that the recordingdevices of the protection and control room cannot access each other directly.

The invention discloses a panoramic camera monitoring device capable of automatically removing dust through gas, and solves the problem that a lens is easily scratched and blurred due to the fact thatthe lens is frequently wiped when the monitor is easily dusted in environments having much dust, such as a factory. The panoramic camera monitoring device in the invention comprises a main frame, a fisheye camera and an infrared camera; the main frame is a vertical oval; the fisheye camera and the infrared camera are staggered and encircled in the middle of the main frame; a connecting part is installed on the top of the main frame; an umbrella cover is installed around the connecting part in a sealed manner; a rotary rod is rotatably installed on the connecting part; a control motor is installed in the top of the main frame; the control motor controls the rotary rod to rotate; a connecting rod is vertically welded on the other end of the rotary rod; an electric push rod is vertically welded below the other end of the connecting rod; an air blower is fixedly installed on the electric push rod; a control centre is arranged in the main frame; and the air blower is connected with the control centre through a conductor wire.