35results about How to "Mitigation mismatch" patented technology

The invention discloses a binocular stereo matching method based on a convolutional neural network. For matching cost calculation, context information is integrated by using dense blocks on the basisof initial features. For matching cost aggregation, a regularization cost amount of a small coding and decoding structure is provided. And for parallax calculation, a differentiable soft argmin operation is executed on the parallax dimension of the cost quantity to obtain an initial parallax. For parallax refinement, the initial parallax is guided to be refined by taking a residual block as a mainpart and taking similarity measurement as an auxiliary part. According to the method, four stages of a stereo matching algorithm are strictly followed, and four steps are integrated into one network,so that the network can be trained end to end. According to the stereo matching method, contextual information is integrated in the feature extraction process, mismatching of pixel points in an ill-conditioned area is effectively relieved, memory occupation and running time in the network training / speculation period are remarkably reduced through small and medium coding and decoding structures inthe regularization process, and parallax prediction precision is improved.

The invention provides a long-term high-temperature oxidation-resistant multi-element composite ceramic coating for a carbon / carbon composite material and preparation and application methods thereof. The long-term high-temperature oxidation-resistant multi-element composite ceramic coating is a composite coating on the surface of the carbon / carbon composite material and sequentially (from inside to outside) comprises a SiC connection layer prepared with an embedding method, a SiC sealing layer prepared with a chemical vapor deposition method and a SiO2-Y2Si2O7-ZrSiO4-Al2SiO5 ceramic layer prepared with a brush coating method and an in-situ oxidation reaction method, wherein the SiC connection layer is combined with a matrix and exerts the function of connecting the matrix, the SiC sealinglayer exerts the function of filling defects of an inner layer, and the outermost SiO2-Y2Si2O7-ZrSiO4-Al2SiO5 ceramic layer serves as an oxygen blocking layer and exerts the function of self-healing and oxygen blocking. The long-term high-temperature oxidation-resistant multi-element composite ceramic coating has the advantages of firmness in combination with the matrix, no run-through crack, long-term high-temperature oxidation resistance, and capability of being completely applied to the preparation of non-direct-ablation heat-resistant structural members around the head and of the wings ofa hypersonic aircraft and the preparation of a material for the tailpipe chamber of an aircraft engine.

The invention relates to a water-based heat insulation paint and a preparation method of the paint; the water-based heat insulation paint is prepared by mixing water, film forming matter, a dispersing agent, an anti-freezing agent, a defoaming agent, a film forming auxiliary agent, a thickening agent, functional filler, composite phase change material and titanium dioxide. The water-based heat insulation paint is coated on the surface of a wall body, and has the effects of heat insulation and building indoor temperature regulation after being cured at the room temperature and completely dried.

The present invention discloses an epitaxial and integrated high-contrast grating external cavity surface emitting laser. A common oxidized vertical cavity surface emitting laser (VCSEL) has polarization indeterminacy or instability phenomenon phenomena caused by anisotropy of oxidization limiting layer materials and anisotropy of active region material gains of the common oxidized vertical cavitysurface emitting laser, and a traditional external cavity line width narrowing method is always complex and low in integration level, and is not helpful for a chip level design. The method for growthof a phase matching layer and a cavity length matching layer is employed to prolong the cavity length of the VCSEL, and a dielectric layer with a low refractive index is formed in one epitaxy mode and is taken as a grating support layer and a grating dielectric layer; and moreover, the grating dielectric layer is etched to form grating microstructures, and therefore, the line width is narrowed while light polarization is achieved.

The invention discloses a microblog query expansion method based on multiple layers. The microblog query expansion method based on the multiple layers is characterized in that keywords are extracted from a PRF (Pseudo Relevance Feedback) layer of a corpus corresponding to original microblog query words and a web layer of an external source to serve as candidate query expansion words, the candidate query expansion words and original microblog query sentences are merged as a label set for labeling documents in the PRF layer, moreover, Labeled LDA is utilized to semantically model for the labeled PRF documents, the candidate query expansion words and the microblog query words coming from the different sources are then mapped to a unified semantic layer, the potential semantics of the candidate query expansion words and the microblog query words are mined, and according to the semantic similarity between the candidate query expansion words and the microblog query words, the candidate query expansion words which are irrelevant to the semantics of the microblog query words are filtered out, so that a new microblog query word is formed for more accurate query and retrieval. Compared with the prior art, the microblog query expansion method based on the multiple layers has the advantages of less query drifts, high retrieval efficiency and high accuracy, and in particular, the microblog query expansion method based on the multiple layers effectively integrates expansion words to achieve an optimal expansion effect, so that query results can meet the real information requirement of users.

The invention discloses a carbon / carbon composite material crucible with a composite ceramic layer and a preparation method thereof. The carbon / carbon composite material crucible is composed of a carbon / carbon composite material crucible base body and the composite ceramic layer attached to the inner surface of the carbon / carbon composite material crucible base body, the thickness of the composite ceramic layer ranges from 1 mm to 5 mm, the composite ceramic layer is composed of, by mass, 50%-70% of silicon carbide, 10%-40% of boron nitride and 10%-20% of silicon. According to the preparation method, a split type preparation method is adopted for the carbon / carbon composite material crucible base body with the composite ceramic layer, a composite ceramic layer blank is prepared firstly, then the composite ceramic layer blank is bonded through a binding agent, then carbonization and siliconizing treatment are conducted, and the carbon / carbon composite crucible base body and the composite ceramic layer become a whole. By means of the preparation method, the composite ceramic layer which is thick enough can be obtained, and therefore erosion of silicon steam to the carbon / carbon composite material crucible base body is effectively restrained.

The invention discloses an audio frequency feature library establishing method and device. The audio frequency feature library establishing method comprises the steps of estimating the noise feature of a recording and playing system; according to the estimated noise feature of the recording and playing system, carrying out imnoise processing on audio frequencies in an audio frequency library; extracting features of the audio frequencies which are subjected to imnoise processing; and an audio frequency feature library is established through the extracted features. According to the technical scheme of the audio frequency feature library provided by the embodiment of the invention, comparing with the training method of a traditional CMI (Contend-based Music Identification) system, the imnoise processing is added in the training phase, the audio frequency feature library which is obtained through the imnoise processing is used to relieve the mismatch phenomenon of training signals and test signals, and the accuracy of an audio frequency identification system can be effectively improved.

The invention relates to a video steganalysis method based on motion vector rate-distortion performance estimation: first, dividing a compression video to be tested into a plurality of frame sets, wherein each frame is composed of continuous video frames, and any video frame only belongs to one frame set; performing steganalysis feature extraction on a frame set including a plurality of motion vectors: for each motion vector in the frame set, obtaining a set composed of the motion vector and an adjacent motion vector, calculating the rate-distortion performance of each motion vector in the set, and extracting preset steganalysis features from the frame set; repeating the steps, and successively performing steganalysis feature extraction on all frame sets of the video to be tested; and employing a classifier based on steganalysis features to perform steganalysis determination on each frame set in the video to be tested. The video steganalysis method can effectively detect a present motion vector domain video steganalysis method.

The invention discloses a supercharged three-phase absorbing type energy storage device and belongs to the fields of refrigeration, heating and energy storage. An existing three-phase absorbing type energy storage device has the problems that needed generating temperature is high, pressure is wasted, generating capacity or absorbing capacity is reduced accordingly and the like. The supercharged three-phase absorbing type energy storage device is characterized in that a reversing valve and a supercharger which can be adjusted in a frequency conversing mode are connected in the existing three-phase absorbing type energy storage device. The device enhances generating effect during an energy storage process, reduces the generating temperature needed by an energy storage machine set, and enhances absorbing effect during an energy releasing process. In addition, heat-supplying temperature can be improved during a heat-supplying process, energy storage density can be improved during a refrigerating process, and different output of refrigerating quantity and heat-supplying quantity can be realized through frequency conversion adjustment. The supercharged three-phase absorbing type energy storage device is applied to the fields like solar energy utilization, industrial waste heat utilization and combined cooling heating and power, greatly eases the problems that a motive power source side and a user side are not matched in space, time and quantity, and has great energy conservation and emission reduction significance.

The invention provides a nitride epitaxial growth method comprising the following steps that a substrate is provided; a first aluminum nitride layer is grown on one surface of the substrate under first temperature and first ammonia gas flow; a second aluminum nitride layer is grown on one surface of the first aluminum nitride layer under second temperature and second ammonia gas flow; a third aluminum nitride layer is grown on one surface of the second aluminum nitride layer under third temperature and third ammonia gas flow, and a nucleation layer is formed by the first aluminum nitride layer, the second aluminum nitride layer and the third aluminum nitride layer, wherein second temperature and third temperature are greater than first temperature, and the first ammonia gas flow and the second ammonia gas flow are greater than the third ammonia gas flow; and a nitride layer is grown on the surface of the nucleation layer. The advantages of the nitride epitaxial growth method are that problems of crystal lattice mismatch and thermal mismatch existing in nitride heteroepitaxy growth can be alleviated, stress can be reduced and dislocation density can be reduced.

The invention discloses an epitaxial integrated dielectric film DBR external cavity surface emitting laser. The laser comprises a P-type metal electrode layer, a passivation layer, an upper distributed Bragg reflector growing periodically and alternately, an oxidation limiting layer with a high aluminum component, an active region, a lower distributed Bragg reflector growing periodically, a GaAs substrate layer, an N-type metal electrode layer, a current limiting oxidation hole, a phase matching layer, a cavity length matching layer, a silicon oxide and silicon nitride dielectric film distributed Bragg reflector and a light emitting hole; and the inherent cavity length of the vertical cavity surface emitting laser is prolonged by adopting a method of growing a phase matching layer and a cavity length matching layer, and the silicon oxide silicon nitride dielectric film distributed Bragg reflector is extended by adopting an inductively coupled plasma enhanced vapor deposition method, so that the purposes of controlling light polarization and narrowing line width are achieved. The threshold value of the semiconductor laser is reduced, and the yield of the device is improved.

The invention provides a non-polar ALGAN-based Schottky UV detector. The non-polar ALGAN-based Schottky UV detector is sequentially provided with a substrate, a low temperature AlN nucleation layer, ahigh temperature AlN buffer layer, an AlN / AlxGa1-xN superlattice structure, an n-type doped n-AlyGa1-yN layer, an n-type doped n-AlzGa1-zN absorption layer, an AlN barrier enhancement layer and a metal film layer from bottom to top, wherein an ohmic electrode is extracted in the n-AlyGa1-yN layer, and o<y<z<x<1. The detector is advantaged in that as the non-polar n-AlyGa1-yN layer has a laterallypolarized electric field directed from an anode to a cathode, photogenerated carriers in the n-AlyGa1-yN layer are accelerated to migrate to the ohmic electrode, photocurrent generation efficiency isgreatly improved, and thereby photoelectric conversion efficiency and responsiveness of the device are improved; the AlN / AlxGa1-xN superlattice structure is inserted, buffer layer barrier height is increased, and crystal quality of the AlGaN epitaxial layer is improved; the AlN barrier enhancement layer is interposed between the n-AlzGa1-zN absorption layer and the metal film layer, height and thickness of a Schottky barrier can effectively be increased, through insertion of the AlN / AlxGa1-xN superlattice structure and the AlN barrier enhancement layer, reducing the dark current of the deviceis facilitated, and the signal-to-noise ratio and stability of the detector are improved.

The invention discloses a carbon fiber reinforced ceramic-based composite material and a preparation method of a pyrolytic carbon interface layer. The composite material comprises a carbon fiber reinforcement body, a ceramic matrix and a pyrolytic carbon interface layer; the pyrolytic carbon interface layer is located between the carbon fiber reinforcement body and the ceramic matrix and wraps thecarbon fiber reinforcement body, and the pyrolytic carbon interface layer is formed by depositing pyrolytic carbon on the surface of the carbon fiber reinforcement body through a chemical vapor deposition process. On one hand, mismatching of physical and chemical properties between the reinforcement carbon fiber and the ceramic matrix in the composite material is relieved, and on the other hand,the carbon fiber is protected from being damaged in subsequent preparation of the ceramic matrix. According to the pyrolytic carbon interface layer prepared by the method, the strength and toughness of the composite material are improved, and meanwhile, the fracture behavior of the material is changed from brittle fracture to ductile fracture.

The invention relates to a multi-microgrid cooperative control method and system, computer equipment and a readable storage medium, each microgrid is connected with a power distribution network through an alternating current line, adjacent microgrids are connected through a direct current line, and each microgrid comprises a power generation device and an electric energy storage device; the methodcomprises the following steps: acquiring power grid data of each microgrid in real time; generating an electric energy transmission instruction according to the power grid data; and performing electric energy transmission between the power distribution network and each microgrid according to the electric energy transmission instruction. The computer equipment comprises a controller and a memory storing computer programs for realizing the steps of the multi-microgrid cooperative control method, and the medium stores the computer programs for realizing the steps of the multi-microgrid cooperative control method. By implementing the invention, the problem of cooperative work after a plurality of microgrids are accessed to a large power grid can be solved.

The invention discloses a method and device for evaluating and calibrating parameters for on-line calculation of thermal efficiency of a cement rotary kiln. The method comprises the following steps: calculating a quality threshold of deviation thermal efficiency parameters during frequent operation of the cement rotary kiln; Standard deviation of thermal efficiency parameters; compare the standard deviation of real-time thermal efficiency parameter calculation with the quality threshold of deviation thermal efficiency parameters, and determine the parameter data for thermal efficiency calculation of cement rotary kiln; correction calculation of secondary air intake into the kiln. The invention not only successfully solves the problem of mismatch of some parameters as the production continues when the thermal efficiency model participates in the actual calculation on site, but also ensures the accuracy of the thermal efficiency model calculation and improves the thermal efficiency in the cement clinker firing process. The accuracy of online calculations.

The invention discloses an evaluation and correction method and device for thermal efficiency online calculation parameters of a rotary cement kiln. The method comprises the following steps that a quality threshold value of a deviation thermal efficiency parameter in the frequent operation process of the rotary cement kiln is calculated; the real-time thermal efficiency parameter standard deviation of the rotary cement kiln is calculated by using a thermal efficiency model; the real-time thermal efficiency parameter calculation standard deviation and the quality threshold value of the deviation thermal efficiency parameter are compared, and parameter data for carrying out thermal efficiency calculation of the rotary cement kiln are determined; and the in-kiln secondary inlet air quantity is corrected and calculated. According to the method and device, the problem that part of parameters are mismatched along with continuous production when the thermal efficiency model participates in field actual calculation is successfully solved, the calculation accuracy of the thermal efficiency model is ensured, and the thermal efficiency online calculation accuracy in the cement clinker sintering process is improved.

The invention discloses an audio frequency feature library establishing method and device. The audio frequency feature library establishing method comprises the steps of estimating the noise feature of a recording and playing system; according to the estimated noise feature of the recording and playing system, carrying out imnoise processing on audio frequencies in an audio frequency library; extracting features of the audio frequencies which are subjected to imnoise processing; and an audio frequency feature library is established through the extracted features. According to the technical scheme of the audio frequency feature library provided by the embodiment of the invention, comparing with the training method of a traditional CMI (Contend-based Music Identification) system, the imnoise processing is added in the training phase, the audio frequency feature library which is obtained through the imnoise processing is used to relieve the mismatch phenomenon of training signals and test signals, and the accuracy of an audio frequency identification system can be effectively improved.

The invention relates to a video steganalysis method based on motion vector rate-distortion performance estimation: first, dividing a compression video to be tested into a plurality of frame sets, wherein each frame is composed of continuous video frames, and any video frame only belongs to one frame set; performing steganalysis feature extraction on a frame set including a plurality of motion vectors: for each motion vector in the frame set, obtaining a set composed of the motion vector and an adjacent motion vector, calculating the rate-distortion performance of each motion vector in the set, and extracting preset steganalysis features from the frame set; repeating the steps, and successively performing steganalysis feature extraction on all frame sets of the video to be tested; and employing a classifier based on steganalysis features to perform steganalysis determination on each frame set in the video to be tested. The video steganalysis method can effectively detect a present motion vector domain video steganalysis method.

The embodiment of the invention provides a sequence labeling method and device. The method comprises the steps of obtaining a sentence to be analyzed; inputting a sentence to be analyzed into the sequence labeling model, and outputting a word segmentation result of the sentence and a labeling result of each word; wherein the sequence annotation model is obtained by performing reinforcement learning-based training based on sample sentences and segmented words and annotation data corresponding to the sample sentences. According to the sequence labeling method and device provided by the embodiment of the invention, word segmentation information of the text is implicitly learned through a reinforcement learning technology, and mismatching between segmented words in Chinese sequence labeling and a target can be relieved.

The invention relates to a video steganalysis method and device based on analysis of the number of CAVLC code words and non-zero DCT coefficients. The method comprises the following steps: dividing a to-be-tested compressed video into a plurality of frame groups; for each frame group, obtaining video steganalysis features by analyzing correlation between the CAVLC entropy coding code words and the number of non-zero DCT coefficients of adjacent 4*4 brightness macro blocks; and carrying out steganography classification judgment on each frame group in the to-be-detected compressed video by using the obtained video steganography analysis characteristics and a classifier. According to the invention, four types of sub-features are extracted, and finally 635-dimensional steganalysis features are obtained. The video steganography method based on the DCT coefficient domain can effectively detect an existing video steganography method based on the DCT coefficient domain at present, can effectively relieve the carrier source mismatch phenomenon in steganography analysis to a certain extent, has a large application range, and can effectively analyze DCT coefficient domain video steganography based on different video coding standards.

The invention discloses a high-nickel ternary single crystal, and the chemical formula of the high-nickel ternary single crystal is LiNi x co y mn z o 2 , the invention also discloses a core-shell structure lithium battery cathode material and a preparation method thereof. In the material disclosed by the invention, the core part increases the nickel content (high-nickel ternary material) to increase the battery capacity, and the shell part (low-nickel ternary material) increases the manganese content to improve the structural stability of the material. The core and shell have a similar lattice structure, which can also alleviate the lattice mismatch phenomenon at the interface between the two. The surface coating layer can improve the interface stability of the high-nickel ternary single crystal, reduce the erosion of the electrolyte, and finally effectively reduce the parasitic reaction on the surface of the high-nickel ternary single crystal, and improve the long-life cycle performance of the material.

The invention discloses a manufacturing method of a membrane material for nitrogen removal of natural gas. The method comprises the following steps: preparing an ABS (Acrylonitrile Butadiene Styrene) resin support body, etching through holes of the ABS resin support body, depositing aluminum oxide on the surfaces of the through holes of the ABS resin support body, forming a porous metal Au transition layer on the surface of the ABS resin support body, and coating an SBS (Styrene Butadiene Styrene) polymer layer on the surface of the ABS resin support body.