73results about How to "Reduce the difference" patented technology

The invention discloses a deep tunnel surrounding rock mechanical parameter inversion method based on a three-dimensional brittle failure zone contour and sparrow optimization algorithm-information vector machine (SSA-IVM) combined optimization algorithm. The engineering technical problem that due to the fact that excavation instantaneous displacement is difficult to monitor, deep tunnel surrounding rock mechanical parameters are difficult to reasonably obtain through displacement back analysis is solved. The method comprises the following steps: firstly, constructing a tunnel FLAC3D numerical simulation model with the same ground stress condition at the occurrence position of a three-dimensional brittle failure zone; secondly, taking an absolute error between the total number of computational grid units in the actually measured brittle failure zone and the total number of computational grid units entering a plastic state in the range of the actually measured brittle failure zone after calculation of the FLAC3D numerical model as an optimization objective function; and then, by taking the tunnel surrounding rock mechanical parameters as optimization variables and taking a target function reaching a global minimum value as a target, performing global optimization by combining a tunnel FLAC3D numerical model and adopting an SSA-IVM joint optimization algorithm, thereby obtaining reasonable tunnel surrounding rock mechanical parameters.

Provided is an optical control element capable of arranging an input portion and an output portion of an optical waveguide on the same side of a substrate in which the optical waveguide is formed, and capable of reducing differences in the configuration of each branch waveguide while minimizing optical path differences between branch waveguides. The optical control element has a substrate (1) having an electro-optic effect, an optical waveguide (2) formed in the substrate, and a control electrode that controls an optical wave propagating in the optical waveguide, characterized in that an input portion (incident light (L1)) and an output portion (emitted light (L2)) of the optical waveguide are formed on the same side of the substrate, the optical waveguide has at least one Mach-Zehnder type optical waveguide portion (MZ) having two branch waveguides (21, 22) branched from one optical waveguide and recombining the two branch waveguides to form one optical waveguide, and the branch waveguides have an even number of return portions (A1, A2).

This invention discloses a low-light scene analysis method based on multimodal feature fusion and clustering, belonging to artificial intelligence technology. It constructs a single-branch feature extraction network based on Transformer; a multimodal feature interaction and fusion module to achieve multimodal feature interaction and fusion; a multimodal fusion feature clustering module to cluster the features after multimodal interaction and fusion, utilizing the semantic and spatial distances between different features to achieve clustering while simultaneously downsampling the features, mitigating the problem of loss of detailed edge information in structured downsampling; and a multi-scale feature aggregation and decoding module to receive feature information from the encoding network and classify each feature pixel according to the semantic distance of the multi-scale features. This invention can fully utilize visible light and thermal image information and can be applied to scene analysis and navigation of unmanned systems in diverse and complex low-light scenes.

The embodiment of the application provides a kind of training method of image encoder, image processing method, electronic equipment, storage medium and program product, it is related to image processing technical field, the training method of image encoder includes: training image is input image encoder, visual word symbol for representing image feature is extracted by the image encoder;Clothing text word symbol is generated for training image;The clothing text word symbol is used to describe the clothing feature of the person in the training image;The model parameter of the image encoder is iteratively trained using a combination loss function;The combination loss function is used to make the visual word symbol of the same person of multiple training images be smaller, and make the visual word symbol and the clothing text word symbol of the same training image be larger difference.It reduces the degree of dependence on visual information by introducing text modal, increases the precision of image encoder to identify the same person under different clothing, improves the precision and robustness of person re-identification.

An electrode apparatus for nerve denervation or modulation includes a main body including a shaft; an electrode unit formed to be drawn out from one end of the shaft and configured to denervate or modulate at least part of nerves on a tube in a body; an electrode guide including a plurality of joint units and a wire connecting the plurality of joint units to each other and configured to guide the electrode unit; and a driving unit located inside the main body and configured to drive the joint units and the wire to protrude from the one end of the shaft. The driving unit drives the joint units in conjunction with the wire to have different displacements.

The application provides a gated multi-modal image fusion method and device, electronic equipment and storage medium. The method comprises: acquiring a plurality of modal images corresponding to a target object; inputting each modal image into a feature extraction module of a multi-modal image fusion model for multi-layer feature extraction processing and differential information compensation processing to obtain a plurality of layer feature maps corresponding to each modal image; inputting the third layer feature map corresponding to each modal image into a feature fusion module in the multi-modal image fusion model for feature fusion to obtain an initial fusion feature map; inputting the multi-layer feature map corresponding to the first modal image, the multi-layer feature map corresponding to the second modal image and the initial fusion feature map into a secondary gating fusion module in the multi-modal image fusion model for multi-layer fusion to obtain a target fusion image. Through secondary fusion, the fusion of different modal images can be realized, and the target fusion image information is more balanced.

The present application relates to a kind of facial expression recognition methods based on attention mechanism and self-distillation, comprising the following steps: step 1, constructing facial expression recognition dataset;Step 2, constructing facial expression recognition model, the model is composed of feature extraction module, adaptive channel attention module, remodeling module and self-distillation network;Step 3, the input facial expression image is preprocessed;Step 4, using the training set image in the above facial expression recognition dataset trains facial expression recognition model;Step 5, using the facial expression recognition model trained to the test set and validation set in dataset are inferred classification.The method designs a kind of reasonable and efficient facial expression recognition scheme, it is well guided model to pay attention to those important feature information by the mode of attention mechanism;By a new self-distillation form, refine compressed network knowledge, improve the robustness of shallow network output feature and reduce the complexity of model in inference stage.

The present application belongs to the technical field of computer vision three-dimensional reconstruction, and specifically relates to a semi-supervised single-view 3D object reconstruction method. A small amount of labeled samples are used to train a neural network, and then the trained neural network is used to generate pseudo labels for unlabeled samples and guide the training of the unlabeled samples. At the same time, the present application proposes that a discriminator scores the quality of the generated pseudo labels, and limits the bias of low-quality pseudo labels on model training. In addition, the present application proposes a prototype shape prior module based on an attention mechanism, which serves as a bridge between images and 3D shapes, reduces the difference between the two modalities, and provides shape prior for the neural network, ensuring that the reconstructed 3D shape is natural. Compared with the current mainstream method in the industry, the present application is more accurate in three-dimensional reconstruction of a single image under the condition that the amount of labeled data is very small, and the generated 3D shape is more realistic and natural.

The application provides a quantum dot microsphere fluorescent immunochromatography test paper for combined detection of ASFV antibodies and nucleic acids and an application thereof, and belongs to the technical field of veterinary biological diagnosis products. The quantum dot microsphere fluorescent immunochromatography test paper for combined detection of ASFV antibodies and nucleic acids comprises a base plate, on the base plate, a sample pad, a combination pad, a nitrocellulose membrane and a water absorption pad are sequentially connected in a head-to-tail mode, from the combination pad to the water absorption pad, the nitrocellulose membrane is sequentially provided with an antibody detection line, a nucleic acid detection line and a quality control line, the antibody detection line is coated with a goat anti-pig IgG-Fc antibody, the nucleic acid detection line is coated with streptavidin, and the quality control line is coated with a rabbit anti-sheep IgG antibody; the combination pad is coated with a quantum dot microsphere labeled recombinant protein P22 and a quantum dot microsphere labeled sheep anti-digoxin antibody. The test paper can rapidly, highly specifically and highly sensitively detect ASFV antibodies and nucleic acids.

The breathing machine comprises a main machine and a humidifying tank, the main machine is provided with an inserting base, the humidifying tank is provided with an inserting section, the inserting section can be in inserting fit with the inserting base in the first direction and can be separated from the inserting base in the second direction opposite to the first direction, and the inserting section is provided with an air outlet adapter pipe; the bayonet socket is provided with a receding groove extending in the first direction, so that when the humidifying tank is matched with the main machine in an inserted mode, the air outlet adapter pipe can enter the receding groove in the first direction. In the process of assembling and disassembling the humidifying tank, the protruding air outlet adapter tube does not need to be disassembled, and the humidifying tank inserted with the air outlet adapter tube can be integrally inserted into the bayonet socket, so that the assembling and disassembling steps of the humidifying tank are simplified, and the disassembling step of the air outlet adapter tube in the process of assembling and disassembling the humidifying tank is omitted. In addition, the risk that internal water leaks from the air outlet due to shaking of the humidifying tank is avoided, the cleaning burden of a user is reduced, and the use experience is improved.

The control method for a drone with multiple degrees of freedom flight modes according to the present invention includes: a fuselage section equipped with a battery, the forward direction of which is set as the x-axis; multiple rotors, four or more arranged around the fuselage section, each rotor having its rotation axis aligned in the z-axis direction; an x-axis tilting mechanism section for tilting the multiple rotors about axes parallel to the x-axis; a y-axis tilting mechanism section for tilting the multiple rotors about axes parallel to the y-axis; a first drive motor section for driving the y-axis tilting mechanism section; a second drive motor section for driving the x-axis tilting mechanism section; and a control section for controlling the first rotor, the second rotor, the third rotor, the fourth rotor, the first drive motor section, and the second drive motor section to realize multiple flight modes.

The invention relates to the technical field of semiconductor optoelectronic devices, in particular to a Micro-LED epitaxial structure forming method, an epitaxial structure and a chip. The Micro-LED epitaxial structure comprises a multi-quantum well layer between an n-type GaN layer and a p-type GaN layer; the multi-quantum well layer comprises 3-10 repeatedly stacked basic units, and each basic unit comprises a quantum well layer and a quantum barrier layer on the quantum well layer; the quantum well layer is an InGaN layer, and the quantum barrier layer is a GaN layer; in the step of forming the multi-quantum well layer, in the direction from the n-type GaN layer to the p-type GaN layer, the incorporation amount of the In component influenced by the process condition when the quantum well layer is formed is gradually reduced. According to the invention, the In component contents in different quantum well layers in the multiple quantum well layers of the red light Micro-LED tend to be consistent, so that the light emitting effect of the red light Micro-LED is improved.

This application discloses a brightness adjustment method, a display device, and a computer-readable storage medium. The brightness adjustment method includes controlling the display panel to enter a partial brightness mode after receiving a target instruction instructing the display panel to enter a partial brightness mode; after the display panel enters the partial brightness mode, if the target brightness value of a second area in the displayed image is less than a preset inflection point brightness value, the brightness value of the second area in the display panel is first adjusted to the inflection point brightness value using a brightness level; then, the brightness of the second area in a first image is adjusted using a first calibration value, and the brightness of the second area in a second image is adjusted to the target brightness value using a second calibration value; wherein the first image and the second image are both displayed images after the display panel enters the partial brightness mode, and the first image is positioned before the second image. This solution can improve the problem of unstable display brightness after the display panel enters the partial brightness mode.

The application discloses a display substrate and a display device. The display substrate comprises a driving back plate, a light absorption layer and a plurality of light emitting components. Light emitted by a plurality of sub-light emitting functional layers passes through a color conversion unit and is emitted as light of different colors. Since the plurality of sub-light emitting functional layers in the light emitting component are arranged in a row in a first direction, the distance of the plurality of sub-light emitting functional layers to the light absorption layer is similar in a second direction intersecting the first direction, and the light emitted by different sub-pixels is similar in the case of being absorbed by the light absorption layer, thereby reducing the difference between light of different colors, and further improving the color deviation problem.

The utility model discloses an axial flow water turbine water guide mechanism which comprises an upper ring seat (1) and a lower ring seat (2), a plurality of guide vanes (3) are arranged between the upper ring seat (1) and the lower ring seat (2), rotating shafts (4) are arranged in the middles of the upper ends and the lower ends of the guide vanes (3), driving rings (5) are connected to the upper ring seat (1) in a rotating mode, servomotors (6) are connected to the driving rings (5), and the servomotors (6) are connected to the driving rings (5). The upper ends of the rotating shafts (4) are connected with the driving ring (5) through a first connecting rod (7) and a second connecting rod (8), and a linkage mechanism connected with the lower ends of all the rotating shafts (4) is arranged at the bottom of the lower ring seat (2); the linkage mechanism comprises a plurality of adapter plates (9), the adapter plates (9) are fixed to the lower ends of the rotating shafts (4) respectively, third connecting rods (10) are arranged between the adjacent adapter plates (9), and the third connecting rods (10) are hinged to the adapter plates (9). The utility model has the advantage of low maintenance frequency.

This invention discloses a thermal management system and a vehicle having the same. The thermal management system includes a compressor and a first heat exchanger. The compressor has an exhaust port and an intake port. The first heat exchanger is adapted to exchange heat with a battery, and its two ends are connected to the exhaust port and the intake port, respectively. The compressor is configured to control the temperature T1 of the refrigerant at the inlet of the first heat exchanger to satisfy T2 < T1 < T3 in a uniform temperature mode, where T2 is the lowest temperature of the battery and T3 is the highest temperature of the battery. The thermal management system of this invention, by controlling the temperature of the refrigerant at the inlet of the first heat exchanger, achieves uniform temperature treatment of the battery, thereby reducing the temperature difference of the battery, resulting in better charging uniformity and extending the battery's service life.

The invention discloses a severe environment unmanned aerial vehicle aerial photography target detection method based on multispectral iteration enhancement, and belongs to the technical field of computer vision. The method comprises the following steps: firstly, acquiring RGB pictures and infrared pictures in severe environments such as rain and fog, night and the like, and constructing a data set; then, an unmanned aerial vehicle aerial photography target detection model is constructed, and efficient deep fusion of visible light and infrared multispectral features is realized by introducing a novel iterative hierarchical attention and differential enhancement fusion framework; meanwhile, a tail enhancement unit of an HGBlock module in the backbone network is redesigned; then training an unmanned aerial vehicle aerial photography target detection model by using the data set; and finally, using the trained detection model to detect an aerial target of the unmanned aerial vehicle and evaluate the performance of the model. Through a fusion mode of combining interactive alignment, differential enhancement and iterative feedback, the target detection performance and real-time performance of the unmanned aerial vehicle under complex weather and illumination conditions are improved.

The utility model relates to the technical field of semiconductors and photovoltaics, in particular to an air exhaust structure and processing equipment, and solves the problem that in the prior art, air exhaust amount difference is large when air in multiple environments is exhausted through one pipeline sharing an air exhaust device. The air exhaust structure comprises an air exhaust pipeline, the air exhaust pipeline is provided with a pipe cavity, the air exhaust pipeline is provided with an air exhaust opening, the pipe wall of the air exhaust pipeline is provided with a first air inlet and a second air inlet, and at least one flow uniformizing piece is arranged in the pipe cavity of the air exhaust pipeline to divide the pipe cavity into a second cavity and at least one first cavity. Each first cavity is communicated with at least one first air inlet, each second cavity is communicated with at least one second air inlet, and suction force can act on air at each first air inlet and air at each second air inlet under the condition that the air extraction opening executes air extraction. According to the air exhaust structure and the processing equipment provided by the invention, the difference of the air displacement at each air inlet in the air exhaust structure can be reduced, so that the air exhaust uniformity and stability at each part are improved.

This application discloses an electrode plate, an electrolysis device, and an electrolytic hydrogen production equipment, relating to the field of electrolysis equipment technology. The electrode plate includes a main electrode plate, an electrode frame, and a distribution box. The electrode frame surrounds and connects to the main electrode plate, forming a receiving groove. The electrode frame has a liquid supply channel and an exhaust channel, and an exhaust hole connecting the exhaust channel and the receiving groove. The distribution box is disposed within the receiving groove and has a flow distribution cavity. The distribution box has a flow distribution surface opposite to the exhaust hole, and the flow distribution surface has multiple distribution holes connecting the flow distribution cavity and the receiving groove. The electrode frame has an inlet hole connecting the liquid supply channel and the flow distribution cavity. The technical solution provided by this application aims to achieve a more uniform distribution and flow of the electrolytic solution within the electrode plate, improving the practicality and reliability of the electrode plate.

A sensing structure for a capacitive accelerometer comprises a resonator structure including a proof mass (101b) and a flexible support (107b) for mounting the resonator structure to a substrate. A gap (dsmall) is defined along the sensing axis (102) between each fixed capacitive electrode (103b) and the proof mass (101b) under zero applied acceleration. The flexible support (107b) has a width (dwidth) along the sensing axis (102) that is substantially equal to the size of the gap (dsmall).

The photoelectric detector comprises a first photoelectric module (3) and a second photoelectric module (4) which are arranged in a stacked mode, the first photoelectric module (3) is provided with an assembling face, the assembling face is provided with a first light sensing part (31), and a transmission filling layer (5) is arranged between the assembling face and the second photoelectric module (4). And the transmission filling layer (5) is used for enhancing the irradiation intensity of light to the first photoelectric module (3). According to the invention, the optical responsivity of the second photoelectric module can be improved, the difference value between the optical responsivity of the second photoelectric module and the optical responsivity of the first photoelectric module can be reduced, and the design complexity and design cost of a rear-end circuit can be reduced.

A rotating electrical machine control device shifts a control system to synchronous five-pulse control when an operating point crosses a second boundary K22 from a state in which asynchronous pulse-width modulation control is being executed, and shifts the control system to the asynchronous pulse-width modulation control when the operating point crosses a first boundary K21 from a state in which the synchronous five-pulse control is being executed. The first boundary K21 is set such that the number of switching pulses per unit rotational speed by the synchronous five-pulse control immediately before the operating point crosses the first boundary K21 is smaller than the number of the switching pulses per unit rotational speed by the asynchronous pulse-width modulation control immediately after the operating point crosses the first boundary K21.

The application discloses an intelligent fire point recognition method fusing a physical optimization model and deep learning, and comprises the following steps: step 1, threshold dynamic processing is performed according to a physical mechanism model to obtain an optimized adaptive threshold physical mechanism model; step 2, a U-Net network model is constructed; step 3, a YOLOv5 model is introduced to detect highlighted targets; step 4, the optimized physical mechanism model and the U-Net network model are used to make a decision to recognize fire points, and remote sensing fire point preliminary screening results are obtained; and step 5, the highlighted targets recognized by the YOLOv5 model are used as error elimination on the remote sensing fire point preliminary screening results, so that the overall precision of the remote sensing fire point preliminary screening result recognition is improved; the problems of fire point false detection and missed detection caused by a fixed threshold are solved, various heat source errors in urban areas that seriously interfere with fire point detection for a long time are eliminated, and the overall precision is further improved; and the problems of poor universality caused by inherent system errors of traditional single thought algorithms are overcome.

The invention relates to a monocyte simulation composition, a preparation method, a quality control substance and a quality control method. The mononuclear cell simulation composition comprises mononuclear cell simulation particles stored in a preservation solution, the volume ratio of the mononuclear cell simulation particles in the composition is 50% or above, the mononuclear cell simulation particles are obtained by treating neutrophils or lymphocytes separated from blood of mammals, and the mononuclear cell simulation particles are used for simulating the mononuclear cells or the lymphocytes. When the mononuclear cell simulation particle is tested by a hematology analyzer, the mononuclear cell simulation particle has the same or similar light scattering characteristics as human mononuclear cells, and the light scattering characteristics at least reflect the complexity of cell contents. The mononuclear cell simulation particles are independently prepared, so that the particles have good consistency and stability, and the obtained composition is small in batch-to-batch difference when serving as a quality control substance, so that the quality control of a hematology analyzer is facilitated, and the reliability of clinical examination of the hematology analyzer is improved.

The invention provides a display panel, a manufacturing method of the display panel and electronic equipment. The display panel comprises a substrate, an isolation structure, a light emitting device and a packaging unit, the packaging unit comprises a first packaging part and a second packaging part, the first packaging part is located in the isolation opening and located on the side, away from the substrate, of the light-emitting device, and the second packaging part is located on the side, away from the substrate, of the isolation structure; the second packaging parts of at least part of the packaging units comprise first sub-parts, and the first sub-parts are located on the sides, away from the substrate, of the second packaging parts of the other adjacent packaging units; the thickness of the first sub-part of the second packaging part is smaller than the maximum thickness of the first packaging part. By reducing the part, covering the other packaging unit, in the packaging unit, the segment difference caused by overlapping of the packaging units can be reduced, the film forming continuity of a subsequent film layer can be improved, and the risk that the film layer formed before is damaged in the subsequent etching step is reduced, so that the pixel packaging effectiveness of the display panel is ensured.

The application discloses a five-dimensional seismic data interpolation method and system based on coordinates, constructs a point-by-point seismic data interpolation model based on a NeRF theory, modifies an MLP framework by combining an additional convolution layer on the basis of the point-by-point interpolation model, constructs a profile-by-profile seismic data interpolation model, introduces kernel norm regularization to construct an objective function, reads original five-dimensional seismic data D obs and inputs the five-dimensional seismic data with missing traces into the network to realize reconstruction of the five-dimensional seismic data. The application eliminates the need for additional labeled data, utilizes unique characteristics of seismic data, enables a convolution network decoder to output data profile by profile, can significantly improve data processing efficiency by 40 times, and increases kernel norm regularization in the objective function, so that the noise resistance and robustness of the model can be effectively improved.

The invention relates to the technical field of rice processing, in particular to an automatic rice processing production system which comprises a silo, a conveying set, a motor, a support, a shell, a sleeve, a first-stage impeller, a second-stage impeller, a third-stage impeller and a discharging plate, the shell comprises a material conveying shaft and a discharging opening, the material conveying shaft is fixedly connected to the inner bottom of the shell, the sleeve is connected to the material conveying shaft, and the first-stage impeller, the second-stage impeller, the third-stage impeller and the discharging opening are arranged in the silo. The first-stage impeller, the second-stage impeller and the third-stage impeller are all installed on the sleeve, the three discharging plates are located between the first-stage impeller and the second-stage impeller, between the second-stage impeller and the third-stage impeller and between the third-stage impeller and the inner bottom of the shell correspondingly, and the discharging openings are located in the downward inclined sides of the discharging plates. A first-stage screen and a second-stage screen are respectively arranged in the centers of the first-stage impeller and the second-stage impeller; according to the centrifugal rice huller, the multiple stages of impellers are arranged to conduct step-by-step hulling on rice with different hulling difficulties, and the effect of improving the hulling uniformity of the centrifugal rice huller is achieved in the mode that the radial length of the centrifugal area of the impellers is shortened.

This application provides a method for manufacturing a semiconductor structure and the semiconductor structure itself. The manufacturing method includes: forming a first intermediate structure and a second intermediate structure; the first intermediate structure includes a first insulating layer having a plurality of first grooves; the second intermediate structure includes a first carrier board and a plurality of first electrical components mounted on one side of the first carrier board; providing a first molding die including a first sub-mold and a second sub-mold, the first sub-mold having a first receiving cavity; placing the first intermediate structure in the first receiving cavity, fixing the second intermediate structure to the second sub-mold, and filling the first grooves with molding compound; closing the first sub-mold and the second sub-mold so that each first electrical component enters a first groove, and the molding compound forms a first molding layer; removing the first sub-mold, the second sub-mold, and the first carrier board; forming a first redistribution structure on one side of the electrode of the first electrical component, the first redistribution structure being electrically connected to the electrode of the first electrical component.

The present application relates to the technical field of fermentation medium, more specifically, it relates to a high-density fermentation medium for streptococcus suis and preparation and application thereof.The medium comprises beef infusion powder 1-5 parts, tryptone 1-10 parts, pancreatic casein peptone 1-10 parts, soybean peptone 1-5 parts, dipotassium hydrogen phosphate 1-5 parts, sodium chloride 1-10 parts, yeast extract 1-10 parts and glucose 1-5 parts.