70results about How to "Guaranteed richness" patented technology

The invention relates to a single-photon polarization self-calibrating and locking method and device based on a phase encoding quantum key distribution (QKD) system and a single-photon polarization control based on phase encoding quantum key distribution system. The polarization control scheme is divided into a calibration process and a locking process which are automatically switched through judging the threshold of a feedback signal. According to the scheme, efficient automatic polarization calibration and locking are realized through the matching of a polarization beam splitter, a multi-degree-of-freedom polarization controller and a genetic algorithm; a polarization feedback circuit device comprises the polarization beam splitter, a polarization controller, a single-photon detector, a signal acquiring and processing module and corresponding control circuits. By using the polarization control method, polarization can be continuously and precisely controlled without reset operation in real time under the condition that the strength of the feedback signal is in the single-photon magnitude, and human intervention is not needed in the whole control process.

The invention discloses a method for preparing a tobacco herbal fragrant substance for cigarette solution of an electronic cigarette. The method comprises the steps that waste and inferior tobaccos are extracted in an ultrasound-enhanced supercritical extraction device for 1.5-2 hours at the temperature of 40-55 DEG C in the presence of an entrainer under the conditions that the ultrasonic frequency is 20KHz, the ultrasonic power is 100W, the flow rate of carbon dioxide of 10-20L / h and the extraction pressure is 15-25 MPa, so as to prepare a mixed solution of carbon dioxide and the tobacco herbal fragrant substance, wherein the entrainer is a mixed solvent of propylene glycol, ethanol and distilled water; the mixed solution is separated by virtue of a separation kettle at temperature of 45- 60 DEG C under the separation pressure of 7-12MPa to obtain the tobacco herbal fragrant substance. The tobacco herbal fragrant substance contains tobacco natural fragrant substance components, nicotine and the like and is applied to the cigarette solution of the electronic cigarette so that the cigarette solution of the electronic cigarette is endowed with the same sense organ characteristic and a satisfaction sense of smoking as common cigarettes. The extraction method is easy to carry out, low in cost, and stable and safe in product quality; the environment pollution is avoided; compared with the conventional supercritical extraction process, the method has the advantage that material consumption and energy consumption are reduced.

The invention discloses an SDN (Software Defined Networking) controller for flow classification based on DPI (Deep Packet Inspection) and a machine learning algorithm. In an existing controller module, a DPI mechanism and a machine learning algorithm are combined by adding a self-designed module to perform an application request classification function of a flow data packet. The self-designed module mainly consists of three parts: a data flow feature building module, a feature extraction and classifier training module, and a classifier-based flow classification module. Firstly, a K-means clustering algorithm is used to construct an information feature base; secondly, a DPI technique is used to extract features to form the training module; and finally, the Naive Bayes algorithm is used to perform detailed flow classification on data packets. The SDN controller for flow classification based on the DPI and the machine learning algorithm, which is disclosed by the invention, is relativelyhigh in classification precision and relatively good in flexibility.

The invention discloses a marine mesoscale vortex classification and recognition method based on deep learning. The method relates to the fields of physical oceans, computer graphic image processing and machine learning. An algorithm mainly comprises model construction, backward processing and model algorithm adjustment. The method comprises the steps of firstly, constructing a model, adopting multiple training strategies to train a mesoscale vortex classification model based on a convolutional neural network algorithm, establishing a vortex classification model, and realizing the efficient classification of mesoscale vortexes; and 2, carrying out backward processing, positioning the high-probability vortex pixels according to the probability density map outputted by the model, carrying out combination and rejection on the repeated vortex images, and recovering the wrongly classified data; and finally, adjusting a model algorithm, adding the error classification data into a training data set to retrain the model, establishing an identification model, and finally determining the position of the sea area where the vortex is located. Practice proves that the method improves the automatic classification and recognition efficiency of the vortex, and expands the application of deep learning in the ocean field.

The invention discloses a phase demodulation device and demodulation method capable of suppressing light-intensity fluctuation noise. The phase demodulation device comprises a fiber interferometer and a digital demodulation device. According to the device and method, an improvement is carried out on a conventional phase generation carrier algorithm; multiplying operation is carried out on a fundamental frequency component and a frequency doubling component of a tested phase, and meanwhile, multiplying operation is carried out on the differential value of the fundamental frequency component and the differential value of the frequency doubling component; then, division operation is carried out on the two products above to remove noise caused by light-intensity fluctuation; and the tested phase is calculated through an integration method. Meanwhile, a noise reduction demodulation algorithm is integrated in FPGA and DSP large-scale high-speed signal processors; and according to the characteristics of high parallel processing capability of an FPGA and excellent floating-point calculation performance of a DSP, differential and corresponding phase calculation are partially realized in the two processors respectively, thereby ensuring real-time performance. The phase demodulation device and method suppress the light-intensity fluctuation noise effectively, improve long-term stability and can be widely applied to the fields of high-precision fiber measurement and fiber sensing and the like.

The invention provides a zero sample target classification method based on pseudo sample feature synthesis. The method comprises the following steps of obtaining the samples and the annotation information of other categories similar to a category to be identified; secondly, obtaining the semantic descriptions of a visible class and a non-visible class by means of network capture and the like, andconverting the class description information into the semantic vectors through a natural semantic processing model; calculating a similarity score between each unseen class and each visible class; constructing a convolutional neural network classification model, wherein the model is divided into a feature extraction part and a classification part; for each unseen class, screening N visible classeswith highest scores according to the similarity scores, randomly selecting the samples, and inputting the samples into a feature extraction network to obtain the feature vectors; combining the feature vectors of the N visible classes according to the similarity score to serve as the feature vectors of the unseen classes; training the classification network by using the unseen feature vectors, sothat the samples of the category can be accurately identified under the condition that there is no available training samples of a certain category of to-be-identified targets.

The invention provides an application of lactobacillus plantarum powder in regulating fat metabolism and intestinal flora of fat rats. The lactobacillus plantarum powder can reduce the in-vivo fat content, increased under induction of high-fat diet, of rats and reduce the body fat rate; the FFA content, increased under induction of high-fat diet, in serum of rats can be remarkably reduced; meanwhile, the TG content in the serum of obese rats can be reduced, and the HDL-C content is increased; the lipid metabolism function of the livers of the obese rats can be improved, and activation of a fatty acid synthesis pathway of the obese rats is inhibited; the condition of intestinal flora disorder of the obese rats can be improved, and the flora richness, flora diversity, flora structure and composition of the obese rats can be maintained; increase of relative abundance of firmicutes is inhibited; the relative abundance of bacteroides is improved; the relative abundance ratio of the firmicutes to the bacteroides in intestinal tracts of rats under the condition of high fat diet is reduced; and the similarity of intestinal flora of the obese rats and healthy rats is increased.

Treating original watermark in binary form, the invention constructs pyramidal watermark with descended resolution layer by layer. Meanwhile, using wavelet transform decomposes multimedia data. Then, according to certain rule, some special wavelet coefficients are selected to embed to watermark. Finally, multimedia data with watermark being embedded into can be obtained through wavelet inverse transform. The invention possesses capability of detecting interpolation including interpolation characteristic and region. The invention also discloses rules of classifying interpolation to distinguish occasional interpolation from vicious interpolation. Thus, vulnerable watermark can resist such as JPEG loss compression etc. content lockout operation. The invention insures reliability of multimedia data released or transferred on network so that it is applicable to authentication possibly.

The invention discloses a text-independent speech emotion recognition method. The method comprises the following steps: acquiring real-time audio data, recognizing the real-time audio data through a trained algorithm model for speech emotion recognition, and determining the emotion state of a speaker, being implemented by performing endpoint detection on the acquired real-time audio data by usinga phoneme alignment processing mode; performing feature value extraction on the real-time audio data after end point detection; and inputting the extracted feature values into an algorithm model, anddetermining the emotional state of the speaker according to the emotional state prediction probability output by the algorithm model. The invention also discloses a text-independent speech emotion recognition device and a model training method. By adopting the scheme disclosed by the invention, the recognition accuracy can be effectively guaranteed.

The embodiment of the invention discloses an image processing method and device, computer equipment and a storage medium. The method comprises: obtaining a to-be-processed low-resolution image; performing feature search in a preset reference pool according to the low-resolution image to obtain a reference feature map, wherein the reference feature map is obtained by carrying out feature extractionon a high-definition image set in the reference pool; and performing feature enhancement on the low-resolution image according to the reference feature map to obtain a high-resolution image. Due to the diversification of the images in the reference pool, the reference feature map contains all possible local features, and high-frequency texture information can be provided for each low-resolution image, so that the richness of the features is ensured, and the memory burden can be reduced. Moreover, the reference feature map is searched according to the low-resolution image, and the selected reference feature map can adaptively shield or enhance various different features, so that the details of the low-resolution image are richer.

The invention discloses a cloud robot system. The system comprises a cloud server platform and an entity robot; the cloud server platform comprises a cloud robot engine module and a cloud programming engine module; the cloud programming engine module carries out signal transmission between the cloud server platform and the cloud robot engine module, and the entity robot carries out signal transmission with the cloud robot engine module in a cloud service manner; the entity robot comprises a plurality of hardware function modules; the hardware function modules follow the hardware function module standard, and due to the hardware function module standard, the hardware function module can carry out registering and adding in a cloud service manner. According to the cloud robot system, the hardware cost of the robot is reduced, the reusability and maintainability of the system are improved, rapid calling and command treatment of the robot under the familiar site are used for conducting the execution action, and richness of intelligent services of the robot is ensured.

The invention belongs to the field of artificial intelligence, and relates to a target object behavior prediction method for data offset and related equipment thereof. The method comprises the steps:acquiring a feature variable, carrying out preprocessing and feature screening, and generating a first feature variable set and a second feature variable set; performing weight assignment on the firstcharacteristic variable set, obtaining a third characteristic variable set according to the second characteristic variable set, and performing weight assignment; performing model training according to the first feature variable set and the corresponding weight to output a training result, and performing model training according to the first feature variable set, the third feature variable set andthe corresponding weight to obtain a training result; and when a comparison result of the two training results meets a preset condition, outputting a second training model to perform behavior prediction. The invention also relates to a blockchain technology, and the behavior prediction result can be stored in a blockchain. According to the scheme, the model risk can be reduced while the model prediction precision is improved.

The invention relates to service recommendation methods and devices. One of the methods includes: acquiring a user identifier and corresponding user instant-scene information; querying a user clusterto which the acquired user identifier belongs, wherein the user cluster is obtained by clustering corresponding user identifiers according to all personal service recommendation manners in a personalservice recommendation manner set of a multi-user source; acquiring cluster-universal service recommendation manners corresponding to the user cluster, wherein the cluster-universal service recommendation manners are generated according to personal service recommendation manners which are in the personal service recommendation manner set and correspond to the user cluster; and recommending serviceinformation, which is in the cluster-universal service recommendation manners and corresponds to a met triggering condition in the cluster-universal service recommendation manners, when the user instant-scene information meets the triggering condition. According to the service recommendation methods and devices provided by the invention, unilateral turning of the recommended service information can be avoided, and richness of the recommended service information is guaranteed.

A previewing system of network TV program comprises data collection unit for carrying out data-sampling of TV program, picture organization unit for processing sampled information to be picture format, cartoon generating unit for preparing picture to be cartoon file and for providing preview, browsing unit for obtaining cartoon file from cartoon generating unit and for playing said cartoon file automatically. Its previewing method is also disclosed.

The invention discloses a zero-sample side-scan sonar image target classification method, and the method comprises the steps: obtaining conventional optical images of the same type according to the target type information of any to-be-recognized side-scan sonar image, carrying out type marking of the conventional optical images to acquire some side-scan sonar images which are easier to obtain anddo not contain to-be-recognized types of targets; using style migration method for the side-scan sonar image with a conventional optical image and a side-scan sonar image as input; generating a simulated side-scan sonar image of a certain specific category, training a deep neural network according to the generated simulated side-scan sonar image data set, and finally, using the deep neural networkobtained by training the simulated side-scan sonar data set to be applied to classification and identification of the side-scan sonar image of the certain specific category. The side-scan sonar imageclassification of a specific category of targets can still be accurately realized without available training samples, and the problem that a recognition network cannot be trained due to the fact thatthe samples cannot be obtained is solved.

The invention provides a workpiece defect detection method and device, and the method comprises the following steps: obtaining defect marking data of a to-be-detected workpiece; enhancing the defect labeling data by adopting an image pixel decomposition and reconstruction algorithm; constructing a defect detection model according to the enhanced defect labeling data; and performing defect detection on the to-be-detected workpiece according to the defect detection model. According to the invention, richness of data used for detection can be ensured, so the performance of a detection model can be ensured, and detection precision and stability are improved.

The invention relates to the technical field of molecular biology and particularly relates to a construction method and an application of a microbial diversity library. According to the novel construction method of the microbial diversity library, the microbial diversity library of a metagenome is constructed by virtue of single chain linkage and a special label primer joint; the scheme is applicable to conventional samples, trace samples, FFPE samples and less than 150 cell groups, and the application range of microbial diversity research is extended; and by introducing the single chain linkage and a special label primer, different target populations fragments can be enriched in a small deviation manner, so that the richness of species and the authenticity of abundance of different species are guaranteed. Besides, the construction method further has the advantages that the initial mass is low, and the same sample can be subjected to multi-species authentication.

The invention discloses a game generating method. The game generating method is applicable to being executed in computing equipment provided with voice input equipment. The game generating method comprises the steps of obtaining a sound signal of a user through the voice input equipment of the computing equipment; obtaining a time-domain oscillogram of the sound signal; converting the time-domainoscillogram into a game reference drawing; generating multiple placeholder points in the game reference drawing, wherein each placeholder point corresponds to one chartlet attribute; utilizing a gameengine to generate a user interface of a game according to the placeholder points in the game reference drawing and the chartlet attributes corresponding to the placeholder points. The invention further discloses corresponding computing equipment and a readable storage medium.

The embodiment of the invention provides a news information pushing method and device and a computer device, and relates to the technical field of artificial intelligence. The news information pushingmethod comprises the steps of if a news information viewing instruction is received, obtaining to-be-processed news information from a news provider, and obtaining target news information from the to-be-processed news information; previewing and semantically identifying the title information of the target news information to obtain a first semantic feature corresponding to the title information;previewing and semantically identifying the content information of the target news information to obtain a second semantic feature corresponding to the content information; comparing the first semantic feature with the second semantic feature to obtain a similarity value, and taking the news information with the similarity value greater than or equal to a first preset threshold value in the targetnews information as to-be-pushed news information; and classifying the news information to be pushed, and pushing the news information to be pushed according to the category information obtained by classification.

The invention relates to a furniture board and the fabrication process thereof, in particular to the board of wood furniture and the fabrication process thereof. The furniture board comprises a plurality of small wood blocks (1) tightly connected with each other through a steel bar (2), wherein holes are formed on the small wood blocks; the steel bar passes through these holes to tightly connect the small wood blocks; the steel bar has screws at both ends; and the screws are connected with nuts (3). The fabrication method comprises following steps: (1) cutting a wood material into small blocks; (2) punching holes at corresponding middle positions of the small wood blocks; (3) passing a steel bar with screws at both ends through the holes to sequentially connect the plurality of small blocks; and (4) applying adhesives at the joints between the small blocks, tightening, screwing nuts on the screws, and leveling. The furniture board has the advantages of unique structure, reduced consumption of wood material, higher strength, and arbitrary shape.

The invention provides a current prediction model training method, a current prediction method and a related device, and the current prediction model training method comprises the steps: detecting a current value when an electromagnet is continuously powered on for many times to unlock a platform door, and obtaining a sample current curve, inputting the sample current curve into a nonlinear network to extract nonlinear features under a time dimension as sample nonlinear features, and inputting the sample current curve into a linear network to extract linear features under the time dimension assample linear features; generating a current value when the electromagnet is powered on next time to unlock the platform door according to the sample nonlinear features and the sample linear features, obtaining a predicted current curve, training a nonlinear network and a linear network into a current prediction model according to the predicted current curve, extracting the nonlinear sample nonlinear characteristics under the time dimension through the nonlinear network, extracting linear features of a linear sample under a time dimension through a linear network, and ensuring the performanceof the current prediction model.

The invention relates to the technical field of waste gas purification, in particular to a compound microbial inoculum and a preparation method and application thereof. The preparation method comprises the following steps: (A) carrying out domestication of an original bacterial liquid; (B) carrying out fermentation of a main strain; (C) carrying out mixed fermentation of auxiliary strains; (D) stirring and mixing the main strain liquid and the auxiliary strain liquid, centrifuging, and taking a supernatant to obtain microbial sludge; (E) stirring and mixing the microbial sludge and a drying protective agent, and drying to obtain the compound bacterial inoculums of petrochemical waste gas treatment. According to the preparation method of the compound microbial inoculum of the invention, strains are collected and domesticated in petroleum pollution sources, the main strain is subjected to single fermentation, the strains are subjected to directional mixed fermentation, the bacteria liquids of single fermentation and mixed fermentation are mixed, centrifugation is carried out to obtain the microbial sludge, and the microbial sludge is mixed with the drying protective agent to preparethe compound bacterial inoculums of petrochemical waste gas treatment. The production cycle is short, and the prepared microbial inoculum has high-efficiency degradation capability.

The invention relates to a micro-ecological preparation for biocontrol on tomato anthracnose and a preparation method of the micro-ecological preparation and belongs to the technical field of anthracnose prevention and treatment. The micro-ecological preparation provided by the invention comprises the following raw materials in parts by weight: 80-100 parts of a composite microorganism bacterial agent, 3-5 parts of a plant methanol extract, 8-10 parts of a salt solution, 3-4 parts of glucose, 0.2-0.5 part of polyaspartic acid, 0.1-0.3 part of heteroauxin, 10-15 parts of fruit juice and 20-30 parts of water. The preparation method of the micro-ecological preparation comprises the following steps: compounding microorganism colonies, extracting plant methanol, preparing the salt solution, andthe like. By adopting the micro-ecological preparation, a biological prevention and treatment method is adopted, antagonism and inhibition functions of microorganisms are utilized, bacteria are treated and inhibited with bacteria, the richness of microorganism colonies on surfaces of tomato surfaces is maintained, environment pollution caused by chemical pesticides is reduced, the plant growth vigor is improved, and the ecological benefits are increased.

The invention discloses an interface bonding strength testing device for a polymer coating material. The device comprises a substrate, wherein a fixed base is mounted on the end surface of the substrate; a substrate fixed plug pin hole is formed in the fixed base from bottom to top; a Brazilian disc test piece guide rail and a stretch plate guide rail are transversely arranged in the substrate from top to bottom; one end, away from the fixed base, of each of the Brazilian disc test piece guide rail and the stretch plate guide rail is of open type design, and an observation window is formed inthe substrate and located above the Brazilian disc test piece guide rail. By means of design of the device, a test piece can be effectively prevented from twisting in a stretching process, accuracy ofa measurement result is guaranteed, besides, the defocus phenomenon during microscopephotographing in a loading process can be prevented by means of the Brazilian disc test piece guide rail, and thedevice can also prevent interface blocking during image shooting of a microscope in an variable-angle experiment, and provides guarantee for richness of the experiment.

The invention discloses a flowerpot having the light supplement function. The flowerpot comprises a flowerpot body (1), the flowerpot body (1) comprises an inner layer (2) and an outer layer (3), the height of the outer layer (3) is greater than that of the inner layer (2), a mounting rail (4) is mounted on the outer wall of the inner layer (2), multiple extending-and-contracting rods (5) are connected to the mounting rail (4) in an equally-spaced mode, each extending-and-contracting rod (5) is connected with a rotating seat (6), and a red light bead (7) and a blue light bead (8) are mounted on each rotating seat (6); multiple round holes (9) are formed in the positions, between the top of the inner layer (2) and the top of the outer layer (3), of an annular plane, and each round hole (9) is covered with transparent glass (10); each rotating seat (6) is connected with a rotating motor (11), and the extending-and-contracting rods (5) are connected with an extending-and-contracting motor (12). The flowerpot has the advantages of being convenient to carry, small in occupied area and easy to operate, and plants absorb light uniformly.

The invention relates to a furniture frame and the fabrication process thereof, in particular to the frame of wood furniture and the fabrication process thereof. The furniture frame comprises one or more small frame segments (1) connected with each other, and a cover (5) at the end head position, wherein each small frame segment (1) includes a plurality of small wood blocks (2) having holes at the middle positions and tightly connected with each other in series through a steel bar (3); the steel bar (3) is provided with screws at the end heads; the screws are connected with nuts (4); the small frame segments are connected with each other through screw threads and with the cover through screw threads. The fabrication method comprises following steps: (1) cleaving a plate material into strips; (2) cutting the strips at a cutting angle of 0 to 45 DEG; (3) cutting the strips into small blocks; (4) punching holes at the middle positions of the strips; (5) passing a steel bar with screws at both ends through the holes to sequentially connect the plurality of small blocks; and (6) applying adhesives at the joints between the small blocks, tightening, screwing nuts on the screws, and leveling.

The invention discloses a potential user screening method and device, a computer device and a storage medium. Firstly, the common user is determined according to the user data of a third party platform and the same user field in the user data of the platform, and then the target user identification of the common user is obtained in the user data of the platform according to the obtained product identification. Then the purchased product information identified by each target user in the user data of the third party platform is obtained according to the target user identification, according to the information of purchased products and a third-party platform with abundant data, a correlation degree calculation formula corresponding to the product identification is constructed, so that the richness of the data can be ensured and the accuracy of subsequent calculation is improved. According to the correlation degree calculation formula, the correlation degree score of each user is calculated in the user data of the third party platform, and the user whose correlation degree score reaches the correlation degree threshold value is obtained, as the potential user, so that the accuracy of the calculation of the potential user is improved.

The invention provides a residual force vector damage identification method based on probability uncertainty. The residual force vector damage identification method comprises the steps of measuring modal information of a local measurement point of a to-be-identified structure; constructing a function equation between the modal information and the structural damage parameters of the to-be-identified structure; obtaining probability distribution of modal information by using a probability uncertainty quantification technology; obtaining a probability distribution solving formula of the structural damage parameters according to the probability distribution of the modal information and the function equation; obtaining damage parameter probability distribution of a damaged structure and an undamaged structure in the to-be-identified structure according to a probability distribution solving formula of the structure damage parameters; and obtaining a damage expectation index of the to-be-identified structure according to the damage parameter probability distribution of the damaged structure and the undamaged structure. According to the method, the damage expectation index representing thestructural damage can be obtained through the modal information of the local measurement points of the structure, good adaptability is achieved, richness of structural damage description can be guaranteed, and meanwhile definite physical significance is achieved.

The embodiment of the invention discloses a text processing method and device, a terminal and a medium. The method comprises the steps of obtaining a candidate response text set corresponding to a target input text; obtaining a first confidence coefficient of each candidate response text in the candidate response text set; wherein the first confidence coefficient of any candidate response text inthe candidate response text set is predicted according to global matching information of any candidate response text; wherein the global matching information is obtained by fusing first matching information and second matching information, the first matching information is used for reflecting the similarity between any candidate response text and the target input text, and the second matching information is used for reflecting the similarity between any candidate response text and another candidate response text; and selecting a target response text from the candidate response text set according to the first confidence of each candidate response text. According to the embodiment of the invention, the target response text corresponding to the input text can be better obtained, and the accuracy of the target response text is improved.

The invention provides a big data analysis system for public equipment management. The system comprises public equipment, first relay terminals, second relay terminals, a processing terminal and a management database, wherein a corresponding jurisdiction region is arranged in each of the first relay terminals and the first relay terminal is connected to public equipment in the jurisdiction region and receives detection information; a corresponding jurisdiction region is arranged in each of the second relay terminals, a second processing policy is used for analyzing the detection information to generate operation information, and the operation information reflects the operating state of the corresponding public equipment; the processing terminal is connected to each of the first and second relay terminals and receives the detection information by means of the first or second relay terminals; and the management database is connected to the processing terminal for storing the detection information.