139results about How to "Make up for the lack" patented technology

The invention relates to a verification method for silt coast wave-induced current numerical simulation, and belongs to the technical field of offshore water dynamic simulation and verification. The verification method for silt coast wave-induced current numerical simulation comprises the following steps of respectively building a Deflt3D wave-induced current calculation model and an ROMS wave-induced current calculation model; respectively calibrating and verifying the applicability of the Deflt3D wave-induced current calculation model and the applicability of the ROMS wave-induced current calculation model; comparing and analyzing commonness and differences of simulations results of the Deflt3D wave-induced current calculation model and the ROMS wave-induced current calculation model, and utilizing a comparison method for verifying calculated results of the models. The verification method has the advantages that on the current condition that silt coast wave-induced current lacks measured data and cannot verify the models, the model comparison verification method is provided, the method is reasonable and feasible, and the defect of lack of measured data of the silt coast wave-induced current can be overcome.

The invention belongs to the field of foods, and particularly relates to a protein and probiotic composite food with the weight losing effect. The protein and probiotic composite food adopts probiotic and protein as main raw materials and is prepared by adding dietary fiber, prebiotics, auxiliary functional factors, vitamins and mineral substances. The protein and probiotic composite food with the weight losing effect contains the probiotic and can change intestinal flora, increase the number of beneficial bacteria, decrease the number of harmful bacteria and decrease the inflammatory reaction, and therefore the probiotic role on a host is achieved; the composition of enteric microorganisms can be adjusted, and therefore the obesity symptoms can be improved. According to the protein and probiotic composite food, the food not only can rapidly and conveniently provide various nutrients for the human body, but also has the advantages of being high in fiber, low in heat, likely to generate satiety and the like and is an ideal and healthy weight losing helper, the body weight is controlled on the basis of balanced nutrients, the healthy weight losing purpose is achieved, obesity complications are prevented, and therefore the food can be specially provided for the obesity.

The invention discloses a traditional Chinese medicine recognition method and system based on GPU and double-scale image feature comparison. The method comprises the following steps: firstly, carryingout dual-scale image acquisition through a specially designed image acquisition device, sending a training image into a convolutional neural network for training, and carrying out multi-layer convolutional pooling feature extraction and selection, and training to obtain a convolutional neural network model with relatively high identification precision; and sending the to-be-detected image into the trained convolutional neural network model for feature extraction, classifying the traditional Chinese medicinal materials based on the extracted features, and outputting a classification recognition result. The method is beneficial to improving the medicinal material recognition capability of ordinary people, and helps experts in the medicinal material field to identify medicinal materials moreaccurately and quickly.

The invention provides an automatic construction implementation method of a software engineering knowledge base based on semi-supervised learning, and solves the problems that the knowledge base of the field of the software engineering is rare at present, the number of concepts is difficult to reach the degree of large scale, the concept-to-concept relation is sparse, and investment of a large amount of labor energy is required. The method is implemented by the steps that firstly, the concept set of the field of the software engineering is expanded by using the Wikipedia and StackOverflow data source according to tag propagation; secondly, machine learning characteristics for performing hyponymy relation extraction on the concepts of the field of the software engineering are constructed; thirdly, positive example and negative example training data of relation extraction are automatically marked by adopting the method of template matching and rule matching; fourthly, the concept-to-concept relation extraction work is performed according to the iterative semi-supervised learning method, and the extraction result of each time of iteration is optimized by combining the evaluation rules; and fifthly, standardized construction of the knowledge base is performed by using the RDF language.

The invention discloses high mixing amount iron tailing sand high intensity concrete and a preparation method thereof. The high mixing amount iron tailing sand high intensity concrete comprises the following components: mixed sand, concrete, coal ash, mineral powder, broken stone, water and a composite additive, wherein the mixed sand adopts iron tailing sand and natural medium sand, the iron tailing sand replaces the natural medium sand in weight percentage, namely the weight percentage of the iron tailing sand in the mixed sand is 60 to 70 percent, and the fineness modulus of the iron tailing sand is 1.9 to 2.6. According to the method provided by the invention, the fineness modulus is selected according to chemical component analysis and grading analysis to the iron tailing sand; the iron tailing sand, the natural medium sand, the cement, the coal ash, the mineral powder, the broken stone, the water and the composite additive are prepared according to working mix proportion. According to the concrete, not only can the performance of a concrete mixture be improved, but also the cohesiveness of the concrete is better, the phenomena of bleeding, separating and the like can be reduced, the high mixing amount of iron tailing sand cannot cause any influences to the intensity of the concrete, an effect of comprehensive utilization to the resource of iron ore tailing can be played, the environment is protected, and the production cost is lowered.

The invention relates to an LED (light-emitting diode) advertisement broadcast monitoring platform and belongs to the advertisement media field and the information technology field. The monitoring platform is composed of a database, a server module, a mobile phone client module and a backstage management system; the database is used for storing all information relating to the advertisement; the server module is used for acquiring relevant data from the database according to the information acquisition request of the mobile phone client, transmitting the data to the mobile phone client, calculating the broadcasting frequency according to the data transmitted by the mobile phone client, and storing the results into the database; the mobile phone client is used for displaying the advertisement information demanded by users, acquiring the advertisement broadcasting data, and transmitting the scanning data to the server; the background management system is used for recording the basic information of all advertisement by administrators, adding up the advertisement broadcasting conditions according to the advertisement broadcasting frequency recorded in the database and generating a report form. In comparison with the existing technology, the whole business in an advertisement putting company is informationized by the monitoring platform; besides, the monitoring platform is high in transmission efficiency and excellent in time validity.

The invention discloses a high-strength concrete and a preparation method thereof. The high-strength concrete is prepared from the following components: mixed sand, cement, coal gangue powder, mineral powder, gravels, water and a compound admixture, wherein the mixed sand is iron tailing sand and natural medium sand, the weight ratio of the iron tailing sand replacing the natural medium sand, namely the percentage of the iron tailing sand taking up the total weight of the mixed sand is 65%-70%, and the fineness modulus of the iron tailing sand is 1.9-2.6. The preparation method comprises the following steps: selecting the fineness modulus according to the analysis on the chemical component and particle size of the iron tailing sand; and preparing the iron tailing sand and the natural medium sand, cement, coal gangue powder, mineral powder, gravels, water and the compound admixture according to the construction mix proportion. By using the concrete, the performance of a concrete mixture can be improved, especially the cohesiveness of the concrete is good, the phenomena of bleeding, dissociation and the like can be reduced, a great amount of iron tailing sand mixed also has no impact on the strength of the concrete and plays a role of comprehensively utilizing iron tailing resources, the environment is protected, and the production cost is lowered.

The invention provides a message processing method and device, equipment and a medium. The method is applied to network equipment, the network equipment is configured with a virtual network adapter driver, the network equipment also comprises a user mode protocol stack and a kernel protocol stack, the method comprises the following steps: creating a virtual network adapter in a system kernel basedon the virtual network adapter driver, sending the virtual network adapter information to the user protocol stack; when a message is received, filtering the message based on the user mode protocol stack; if the message is a control message, by the user protocol stack, forwarding the control message to the kernel protocol stack for processing through the virtual network adapter based on the virtual network adapter information; and if the message is a service message, by the user protocol stack, processing the service message. By applying the method disclosed by the invention, the network adapter can process the service message at high performance in the user mode protocol stack, and the network adapter can support the control plane protocol through the linux kernel protocol stack.

The invention discloses a method for identifying traffic information in social media. The method comprises the following steps of: 1, preprocessing social media data; 2, classifying the social media data; and 3, carrying out named entity identification on the social media data. According to the method, machine learning technologies such as a polynomial model, a conditional random field algorithm and the like are utilized to carry out classification and named entity identification on social media information, so as to extract information related to traffic from the social media and then extractlocations of incidents in the information. The method is capable of timely providing traffic public opinions and information such as situations, influence ranges, origins and the like of emergent traffic incidents for the traffic management department. In regions, the traffic information acquisition and construction of which are relatively weak, the method is capable of providing information supplementation for the traffic management.

The invention discloses a traditional Chinese medicinal material recognition method and system based on a double-scale convolutional neural network. The method comprises the following steps: firstly,carrying out dual-scale image acquisition through a specially designed image acquisition device, sending a training image into a convolutional neural network for training, and carrying out multi-layerconvolutional pooling feature extraction and selection, and training to obtain a convolutional neural network model with relatively high identification precision; and sending the to-be-detected imageinto the trained convolutional neural network model for feature extraction, classifying the traditional Chinese medicinal materials based on the extracted features, and outputting a classification recognition result. The method is beneficial to improving the medicinal material recognition capability of ordinary people, and helps experts in the medicinal material field to identify medicinal materials more accurately and quickly.

The invention discloses a rotating surface based goaf laser scanning point cloud triangulation method. The method comprises steps as follows: 1), irregular actually measured scanning point cloud is converted into spherical projection point clouds round and round through spherical projection and subjected to interpolation, original topological relation is restored, and actually measured scanning point cloud is subjected to interpolation synchronously; 2), points on an upper circle of a rotating surface and a central rotating shaft form the rotating surface, a point, nearest to the rotating surface, of an adjacent circle is searched, and the optimum grid is established; and 3), the optimum grid is subjected to triangulation with a middle position judgment method, a goaf in-situ point cloud is subjected to triangulation synchronously, and a triangular grid model of a goaf is acquired. The method is high in efficiency and accuracy.

The invention relates to a maximum-particle-size structure descriptor-based pedestrian re-identification method. The method comprises the following steps of S1, obtaining colorful pedestrian images in an image set, and processing the pedestrian images by using a Gabor filter to obtain a plurality of dimensional images; S2, obtaining color difference histograms (CDH) of the dimensional images, and extracting LMCC descriptors; S3, extracting LOMO descriptors; S4, performing measure learning by using an LDA algorithm to obtain optimal sub-space projection of a feature space; and S5, inputting to-be-identified pedestrian images, and calculating similarity measure distances between the to-be-identified pedestrian images and the pedestrian images in the image set to obtain an identification result. Compared with the prior art, the method has relatively good robustness for change of factors such as illumination, rotation, translation and the like, can be used for extracting essential features of the images, has good pedestrian identification performance, and is not sensitive to change of illumination, view angles, shielding and the like.

The invention provides a plateau plant solid ferment food. The preparation process of the plateau plant solid ferment food consists of: firstly subjecting Chinese wolfberry fruit pulp, sea buckthorn fruit pulp, Nitraria fruit pulp, a Coreopsis tinctoria extracted liquid, a rhodiola root extracted liquid and a Maca extracted liquid to submerged fermentation to obtain a fermented supernatant, then mixing the fermented supernatant with soybean whey protein powder, and performing drying to obtain solid ferment. The solid ferment provided by the invention takes the plateau plants with very strong adverse resistance ability in the aspect of pharmacology as the raw materials, and employs a compound ratio, has comprehensive and mutually complementary nutrition, and reach collaborative efficacy. Through specific technological treatment on the plateau plants and phased fermentation effect of a variety of probiotics, the nutritional value and health care functions of the raw materials themselves are greatly enhanced, and the plateau plant solid ferment product with the advantages of rich nutrition, high content of active ingredients, good flavor and excellent taste can be obtained.

The invention discloses a preparation method for preparing the lithium-rich composite cathode material of a lithium ion super capacitor. The method comprises the steps of dissolving a soluble nickel manganese cobalt salt in the deionized water in a certain proportion to obtain a mixed solution; dispersing the carbon material high in specific surface area in the above mixed solution; adopting the sodium hydroxide solution and the ammonia water as a precipitating agent and a complexing agent respectively; preparing a mixed suspending liquid of the nickel manganese cobalt salt according to the precipitation method and loading the mixed suspending liquid on the carbon surface and inside holes; filtering and washing with water to obtain an activated carbon-loaded precursor; mixing the precursor with a lithium source, and conducting the solid-phase sintering operation in a tube furnace in a specific atmosphere to obtain the target material. According to the technical scheme of the invention, under the dimensional limitations of porous carbon particles, the lithium-rich material is prevented from agglomerating and the crystal is prevented from growing excessively during the sintering process. Therefore, the material rate capability is improved. Meanwhile, during the preparation process, a lithium source of a super stoichiometric ratio is added, and the part of the lithium source deposits on the surface of the porous carbon material. Therefore, the shortage of the lithium source, caused by an SEI film and the competitive adsorption phenomenon during the first charging process, is compensated.

The invention discloses a sour plum black tea beverage. The sour plum black tea beverage comprises compound type sour plum black tea essence, purified water and auxiliary materials; the compound type sour plum black tea essence consists of sour plum essence and black tea essence; the sour plum essence and the black tea essence are in the mass ratio of (3:2)-(1:1); the sour plum black tea beverage comprises the sour plum essence in the mass percentage concentration of 0.05 to 0.1 percent, the auxiliary materials with the mass percentage concentration of 6 to 6.5 percent, and the balance of the purified water. The invention also provides a preparation method for the sour plum black tea beverage. The sour plum black tea beverage is prepared by compounding the sour plum essence and the black tea essence in a certain ratio, adding the purified water and suitable auxiliary materials in sequence, homogenizing, and pasteuring and filling. According to the sour plum black tea beverage, the compound type sour plum black tea essence is compatible with a natural extract to form a mixture; the mixture serves as a flavour component and is applied to a beverage formula, so that a flavour substance achieves an effect of providing the aroma of fresh sour plum and Guangdong Yingde black tea in the beverage; the characteristic aroma lost after a heat processing treatment of concentrated juice and the black tea is made up.

The invention provides a Cocone curved surface reconstruction method for maintaining object surface sample point edge characteristics. The method is characterized in that edge characteristic reconstruction is optimized by correcting a normal estimation result and adjusting the Cocone angle threshold value of edge characteristic sample points. Gain optimization is performed on a localized curved surface sample used by PCA normal estimation so that the normal estimation result is corrected. An adjusting formula for the Cocone angle threshold value of the edge characteristic sample points is established so that the size of the angle threshold value of the edge characteristic sample points is adaptive to sample point distribution of an edge characteristic region. In the angle threshold value adjusting process, identification of the edge characteristic sample points is realized on the basis of Gauss clustering of the localized curved surface sample of the target sample points, and a cross-region long and narrow surface patch generated by adjustment of the angle threshold value of the edge sample points can be rapidly filtered on the basis of the predetermined scale threshold. Compared with methods in the prior art, defects of the reconstruction result of recesses and holes in the edge characteristic region can be substantially reduced.

The invention belongs to the field of medical technology, and discloses an atrial fibrillation patient information management and service system based on Internet and a method. The atrial fibrillationpatient information management and service system comprises the components of an application layer, an official account service management terminal and a database server. The application layer is thefirst layer and is used for supporting a PC computer terminal and mobile equipment. A patient and medical personnel enters a medical communication system through a Wechat official account platform. The system mainly performs data browsing, inquiring, recording of a small number of texts or image uploading in modes such as online H5 PAGE. The official account service management terminal is a second layer and is used for receiving a client request, performing exchange with a background server, generating information according to service logic and transmitting the information to the H5 page fordisplaying. The database server is a third layer and is used for organizing and storing data such as patient information, medicine information, treatment result information and knowledge database.

The invention provides an experimental device for gas leakage and diffusion of an underground pipe gallery. The experimental device comprises a gas cabin of the underground pipe gallery, a digital display system, a ventilation system, a gas supply system, and a tail gas treatment device. One end of the top surface of the gas cabin of the underground pipe gallery is connected with the ventilation system, and the bottom side of the end of the gas cabin of the underground pipe gallery is connected with the gas supply system. The digital display system comprises a display, a synchronous controller, and a particle image velocimeter. The particle image velocimeter is arranged outside the cabin body of the gas cabin of the underground pipe gallery. The particle image velocimeter, the ventilationsystem and the gas supply system are respectively connected with the synchronous controller. The experimental device of the invention can be used for researching and analyzing the dynamic change ruleof the gas leakage and diffusion flow field of the gas cabin of the underground pipe gallery, and has important application value on the safety and control for the gas leakage accident of the underground pipe gallery.

The invention discloses a method for precisely measuring an ultraviolet absorbent in a household product and a pretreatment system. The method comprises the following steps: placing a household coating product in an extraction tube of an extraction and purification integrated test tube, adding ethyl acetate for dispersion, extracting by using an ammoniation organic reagent, obtaining an extractingsolution by vortex, ultrasonic and centrifugation, purifying by using a purification column, concentrating by using a high-efficiency concentration and enrichment device, dissolving residues by usinga formic acid and methanol-acetonitrile mixed solution, carrying out vortex, filtering, collecting filtrate to obtain a solution to be measured, and detecting by using a liquid chromatography- tandemmass spectrometer. The method has simple and rapid pretreatment, solves the problem of inaccurate qualitative and quantitative determination of the ultraviolet absorbent of the household coating product by adopting a matrix matching method, is suitable for detecting the household coating product, and can also be used as a reference method for detecting the ultraviolet absorbent in other householdproducts. The invention also discloses a pretreatment system, which comprises an extraction and purification integrated test tube and an efficient concentration and enrichment device.

The invention discloses a hyperspectral image depth noise reduction method and system based on a two-stage learning framework, and belongs to the technical field of hyperspectral image noise reduction, and a three-dimensional depth universal model 3D-DUSSD for hyperspectral image noise reduction comprises the following steps: constructing a target function based on noise estimation and image noise reduction; constructing a conditional estimation sub-network (CENet) and a multi-scale cross fusion noise reduction sub-network based on a target function; training a conditional estimation sub-network (CENet) and a multi-scale cross fusion noise reduction sub-network by using an objective function; the trained conditional estimation sub-network (CENet) is utilized to deduce the noise level of the hyperspectral image, and then information is transmitted into the multi-scale cross fusion noise reduction sub-network to carry out the denoising of the hyperspectral image. By means of the method, non-i. I. D. Noise distribution can be predicted, and meanwhile deep noise reduction of the hyperspectral image is achieved.

The embodiment of the invention provides an intranet user abnormal behavior recognition method and device based on density clustering, computer equipment and a computer readable storage medium. The embodiment of the invention belongs to the technical field of safety monitoring. The method includes: obtaining user data of an intranet user, wherein the user data comprises a network access log and attendance data; according to the user data, describing behavior portraits of the intranet users by taking a preset time period as a time slice; obtaining a plurality of feature data sets of the time sequence; clustering the feature data set through a preset density clustering model; obtaining clustering results, judging whether outliers exist in a clustering result or not; if the outliers exist inthe clustering result, determining that the outliers exist. Compared with safety monitoring of daily behaviors of the intranet users through entity monitoring rules in the prior art, the method and device have the advantages that the recognition efficiency, accuracy and comprehensiveness of the abnormal behaviors of the intranet users can be improved, and risk behaviors of the intranet users in enterprises are avoided.

The present invention belongs to the field of luminescent materials, and discloses a manganese-doped aluminosilicate oxynitride luminescent film, a preparation method and organic electroluminescent devices thereof. The manganese-doped aluminosilicate oxynitride luminescent film comprises the following chemical components, by mass, 0.05-5% of Al2O3, 0.01-1.5% of MnO2, and the balance of Si3N4. According to the present invention, magnetron sputtering equipment is adopted to prepare the manganese-doped aluminosilicate oxynitride luminescent film, wherein the film does not contain noble metal rare earth and other elements, and has characteristics of simple preparation and low cost. In addition, in a luminescence spectrum of the film of the present invention, luminescence in a blue light region can be achieved, and the whole blue light wavelength range of 490-550 nm are covered, such that blue light material deficiency in LED display and lighting can be overcome.

The invention discloses a three-dimensional hollowed-out enamel handicraft. The handicraft comprises a three-dimensional frame, wherein more than one hollowed-out pane is arranged on the surface of the three-dimensional frame, at least one hollowed-out pane is an arc-surface hollowed-out pane, and the hollowed-out pane is filled with enamel glaze through a hollowed-out enamel process. The handicraft has the characteristics of three-dimensional effect, aesthetic feeling, and being more prominent in clearness and ambilight of hollowed-out enamel, more ornamental angles can be brought, more gorgeous impression experience can be provided, through the hollowed-out enamel effect based on cambered surfaces on the handicraft, the curved aesthetic feeling brought by the cambered surface can be increased, and the overall stereoscopic impression can be highlighted. By combining patterns, and the shape of the pane, more diversified authoring spaces based on the three-dimensional hollowed-out enamel handicraft can be provided, and the development of the handicraft made based on the hollowed-out enamel process is facilitated.

The invention belongs to the technical field of acid soil improvement and specifically relates to a straw decomposition promoting method and application thereof to acid soil improvement. According tothe method, straw is utilized as a raw material, and calcium salt and sepiolite powder are utilized as decomposition promoting aids; the two aids of the calcium salt and the sepiolite powder are combined, so that a straw decomposition process can be obviously quickened, and straw decomposition time is obviously shortened; meanwhile, mixed decomposition of the straw with human and livestock excrement is avoided, and environmental friendliness of a decomposition process is achieved; the calcium salt in the aids can improve a calcium content in a decomposition product, the sepiolite powder can provide a good place and a good environment for EM bacteria growth and propagation, a humic acid content in the decomposition product is increased, complexity is high, more metal ions can be combined, and soil base ion retention is facilitated. When the obtained decomposition product is applied to acid soil improvement, the pH of the soil can be obviously improved, a content of exchangeable aluminumis reduced, a content of soil microbial biomass carbon is further increased, and carbon metabolic activity of microorganisms is improved.

The invention discloses an entity disambiguation method based on a neural network combined with knowledge description, and relates to the technical field of natural language processing. The method comprises the following steps: 1, modeling by using a referring context text and a candidate entity, and calculating the similarity between the referring context text and the candidate entity; 2, modeling by using text information described by candidate entity knowledge and referred context text; 3, extracting candidate entity knowledge description keywords; 4, establishing a local model of entity disambiguation; 5, establishing a global model of entity disambiguation; and 6, introducing a loss function, and training to find the target formula in the step 4. According to the short text entity disambiguation method, entity operation is carried out simultaneously from the candidate entities and the vectors of the context, the global model is used for entity disambiguation of the short text, and the optimized local model is combined, so that the problems that a global model corpus lacks and the local model lacks other entity information of the text are solved.

The invention discloses a video face recognition method combining deep Q learning and an attention model. The method comprises the five steps of video feature extraction, video time continuity information extraction, local face positioning, video optimal frame sequence sorting and video face recognition matching, and has the advantages that the video face recognition can be achieved, and the problem that the video face matching precision is insufficient is solved.

The invention belongs to the technical field of agricultural resource, and particularly relates to a production system and a production method of a carbon-based organic-inorganic compound fertilizer,wherein the production system comprises a pretreatment device, a carbonization furnace, an oil-gas separation system, an oil-water separation system, a fermentation system, a drying and crushing system, a mixing granulation system and a steam boiler system. According to the present invention, the oil-gas resources generated in the carbonization process are recovered and utilized, and the biochar produced by pyrolysis and the chemical fertilizer are combined to produce the carbon-based organic-inorganic compound fertilizer, such that the returning of the agriculture and forestry biomass straw to the soil is achieved, the organic matter of the soil is increased, the adsorption performance of biochar is utilized, and the slow release of chemical fertilizer nutrients is achieved so as to improve the utilization efficiency of chemical fertilizers.

The invention relates to the field of livestock and poultry feed and in particular relates to feed with a tea flavor special for cage broiler. The feed is prepared from the following raw materials in percent by weight: m35-45 parts of barley malt root, 8-10 parts of sweet potato leaf, 15-20 parts of rice sugar residues, 150-180 parts of corn, 35-40 parts of sorghum wheat middling, 45-50 parts of rye bran, 35-40 parts of mung bean straws, 2-3 parts of pig bone powder, 1-2 parts of calcium hydrophosphate, 4-5 parts of Swiss chard leaf, 1-2 parts of swine blood meal, 2-3 parts of fly maggot meal, 2-3 parts of black tea, 2-3 parts of green tea, 2-3 parts of tea flower, 2-3 parts of dried tea seed powder, 3-4 parts of perilla leaf, a proper amount of salt and 4-5 parts of phagostimulant. The broiler feed provided in the invention contains the black tea, green tea, tea flower and tea seeds, contains rich proteins, also contains multiple trace elements and biological active ingredients, the loss of nutrient elements in the traditional cage broiler feed is avoided, all the raw materials in the feed are pure natural green components, the fed broiler is delicate and delicious in meat, unique in flavor and high in edibleness.

The invention discloses a method for generating an automatic test case of a rail transit signal system. The method comprises the following steps of S1, performing case scene analysis; S2, extracting specific influence factors of the scene; S3, orderly combining the influence factors through a CT algorithm, and automatically generating test data covering all scenes and all function points and faultpoints; S4, automatically screening and eliminating the test data; S5, selecting a part of test data results for manual analysis to obtain a data set with labels, carrying out modeling analysis on the obtained data set by using an LSSVM algorithm to obtain an LSSVM model, judging and identifying an expected result of an unknown sample by utilizing the LSSVM model to obtain the expected result ofeach piece of sample data, and combining the expected result with the corresponding sample data to obtain a finished test case. According to the invention, automatic generation of the test cases is realized, and the test cases are completely covered without omission.

The invention discloses a packaging structure of a silicon carbide Vienna rectifier half-bridge module, and belongs to the technical field of power semiconductor devices. The packaging structure comprises a power unit and an insulating substrate. The power unit comprises a diode rectification upper bridge arm, a diode rectification lower bridge arm and a bidirectional switch bridge arm. The bidirectional switch bridge arm comprises a first silicon carbide MOSFET chip and a second silicon carbide MOSFET chip which share a common source and are connected in series. The surface copper layer of the insulating substrate comprises a positive electrode metal layer, a negative electrode metal layer, an input metal layer, an output metal layer, a bidirectional switch power common source electrode metal layer, a driving metal layer and a thermistor metal layer. The diode rectification upper bridge arm is located on the positive electrode metal layer, the diode rectification lower bridge arm andthe first silicon carbide MOSFET chip are located on the input metal layer, and the second silicon carbide MOSFET chip is located on the output metal layer. According to the present invention, the parasitic inductance in a current conversion loop is reduced, so that switching loss is reduced, and the packaging structure is more suitable for high-frequency application; and the power switch tube chip is driven by adopting Kelvin connection, so that the driving stability is enhanced.