553results about How to "Reduce processing" patented technology

The invention provides a network node classification method based on a multilayer graph neural network model, and the method comprises the steps: constructing a graph according to network nodes, carrying out the preprocessing of initial node features in the graph, and obtaining an initial embedded vector; according to the initial embedding vectors of the node and the neighbor node thereof, updating the node representation by adopting a layered message passing mechanism; according to the initial embedded vectors of the nodes and the neighbor nodes thereof, the graph attention network uses an attention mechanism to calculate weights for the neighbor nodes of each node and then dynamically updates feature vectors of the nodes; a multi-head attention mechanism is adopted to splice and merge the feature vectors of the nodes output by the graph attention network and the updated node representation, and the comprehensive feature vector of each node is obtained; the comprehensive feature vectors are further optimized through graph contrast learning and a node embedding module, and optimized features are obtained; and obtaining predicted node classification according to the optimization features. According to the method, complex features of heterogeneous graph data can be effectively captured, the method has good robustness for an over-smoothing problem, more accurate heterogeneous network representation learning is realized, and the method has advanced performance and robustness in a node classification task.

The invention discloses a multi-source ship target identification method based on two-stage collaborative fusion, and relates to the technical field of signal processing, and the method comprises the steps: firstly collecting target sea area data, including original optical image and radar image data, by means of an optical camera and a radar sensor on an aircraft, and then carrying out the preprocessing of the collected data through Gaussian filtering; performing sea-land segmentation by using a deep learning algorithm, constructing a loss function optimization network parameter to obtain a segmentation mask, and removing a land background area according to a mask result to obtain processed optical and radar images; an improved SURF feature and graph cut algorithm combined registration method is adopted, optimal transformation parameters are obtained by screening feature point pairs and solving the minimum value of an energy function, and image registration is achieved; for the registered images, fusing the images based on an adaptive fusion rule; and finally, extracting corresponding features through manual feature extraction and feature extraction based on deep learning, and fusing the features to obtain a final target recognition result.

The invention discloses a brake spring piece flatness detection system and method, and relates to the technical field of automobile part processing. The brake spring piece flatness detection system comprises a data acquisition module, a data analysis module, a data screening module, a dynamic monitoring module and a comprehensive judgment module, wherein the data acquisition module is used for acquiring static detection data of N groups of spring pieces randomly extracted from the same batch; the static detection data comprises top view shooting picture data and side view shooting picture data of N groups of spring pieces, which are acquired at equal intervals by utilizing an optical detection instrument, and size data of the spring pieces, which are acquired by utilizing a laser radar, and the technical key points are as follows: through analysis and processing of the static detection data and the dynamic detection data, the overall quality of the spring pieces is shown; related workers can conveniently and quickly know whether the spring pieces meet the requirements of automobile braking or not, high-quality products can be quickly screened out from numerous spring pieces, use is convenient, secondary detection is not needed in the follow-up process, and good application prospects are achieved.

The invention discloses an intelligent number asking method based on a large language model, and relates to the field of artificial intelligence application, a question proposed by a user is analyzed and understood through an LLM, a number asking API matched with the question is searched according to the question of the user, then a system inputs the searched API, query parameters of the API and the question of the user into the LLM together, and the query parameters of the API and the question of the user are input into the LLM; the LLM further analyzes user questions and API query parameters, then generates an API call request body and provides the API call request body to the system; when the number query API cannot be found, performing a query task by using a query SQL; the problem that the length of the cue word exceeds the context length limit of the LLM due to the fact that the number of APIs for query access is too large and API description is too long is effectively solved, and the situation that when the LLM is pre-trained, some knowledge is not learned or expired, and consequently the problem cannot be solved or mistakenly solved when the number asking requirement of a user is processed is avoided by adding a special label into the cue word.

The invention provides a method for quickly training an LSTM (Long Short Term Memory) model by time series data, which relates to the technical field of data processing, and comprises the following steps: S1, a data integration stage: efficiently processing and utilizing telemetering real-time stream data through Kafka and Flink, s2, a data storage stage: storing the time sequence data processed in the integration stage into a high-performance database IoTDB specially designed for the time sequence data according to a time sequence, and S3, an algorithm application stage: constructing an LSTM model, setting parameters, optimizing the LSTM model, and performing time sequence data prediction by using an LSTM algorithm model. According to the method for rapidly training the LSTM model through the time series data, the time series telemetry data generated in the test flight process are collected and processed through Kafka, real-time data of an aircraft can be rapidly captured, processed and analyzed, data transmitted by a sensor and an algorithm model are decoupled through the Kafka, data output by the sensor can be directly sent to the Kafka, and the LSTM model can be rapidly trained through the Kafka. And it is avoided that an algorithm model needs to be concerned to process data.

The invention relates to the technical field of power grid interaction, in particular to a two-way energy control method and intelligent equipment, and aims to solve the problem of how to perform charging and discharging control on two-way energy equipment according to power grid load information and peak and valley electricity prices on the basis of an optimal cost principle. The method comprises the steps that the initial electric quantity of an energy storage module is acquired; acquiring historical electrical load information of a set time period; and predicting a power control curve of a set time period according to the initial electric quantity, the historical electric load information and the expected electric quantity based on a cost optimal principle. On the basis of the predicted power control curve, charging and discharging control is performed on the energy storage module through the bidirectional energy equipment, so that the stability of the power grid can be enhanced, the power supply condition of the power grid can be optimized, the electric charge expenditure of the user can be reduced, the user experience can be improved, income can be brought to the user, and the return on investment of the user can be improved. And by adopting a dynamic programming algorithm, the operation speed is improved, and engineering realization is more convenient.

The invention relates to the technical field of information recommendation, in particular to an AI-driven customer segmentation and personalized recommendation system which comprises a user acquisition module, a data classification module, a user division module, a training module, a preference link generation module, a retrieval link generation module, a push output module, a monitoring verification module and an analysis module. According to the method, users are subdivided into different groups and divided into a plurality of user categories, personalized link recommendation is provided for each user category, accurate user subdivision and personalized recommendation are realized by collecting and analyzing a large amount of user data, and the data processing efficiency is improved.

The invention relates to the technical field of preparation of lithium battery materials, and discloses a synthesis method of a low-cost battery-grade lithium dihydrogen phosphate raw material lithium phosphate. The synthesis method comprises the following steps: dissolving crude lithium sulfate in alkaline lithium phosphate lithium precipitation mother liquor; carrying out chemical precipitation on the filtered lithium sulfate solution to preliminarily remove calcium, and subsequently, deeply removing calcium and COD (Chemical Oxygen Demand) by sequentially using activated carbon and calcium removal resin to obtain a lithium sulfate finished solution; mixing the lithium sulfate completion liquid and the alkali liquid, freezing to-10 DEG C to-3 DEG C, and performing centrifugal separation to obtain a lithium sulfate freezing liquid; and reacting the refrigerating fluid with a dilute phosphoric acid solution to precipitate lithium phosphate, stirring, washing and centrifuging the filtered crude product lithium phosphate, and drying to obtain the battery-grade lithium dihydrogen phosphate raw material lithium phosphate. The lithium phosphate prepared at low cost is used for replacing battery-grade lithium carbonate, lithium hydroxide and the like as a lithium source for producing the lithium dihydrogen phosphate, so that the production cost of the battery-grade lithium dihydrogen phosphate can be effectively reduced, the lithium phosphate mother liquor is recycled, and the problem of treating a large amount of wastewater caused by other methods is solved.

The invention provides a digital twin modeling method, device and equipment and a storage medium, and relates to the technical field of digital twin modeling, and the method comprises the steps: obtaining initial point cloud data; obtaining preprocessed point cloud data; obtaining first point cloud feature information according to the preprocessed point cloud data; carrying out object classification and segmentation processing to obtain second point cloud feature information; generating an interactive twin model; and acquiring real-time state information, and updating the interactive twinborn model according to the real-time state information. Convolutional feature extraction processing based on point cloud data is realized by generating the point cloud image, global and local features of the point cloud can be efficiently extracted, the modeling efficiency and precision are improved, the high modeling precision is beneficial to guaranteeing the precision of an interactive twinborn model to accurately reflect an actual scene, and the method is suitable for large-scale popularization and application. And meanwhile, the high modeling efficiency is beneficial to realizing synchronous change of the interactive twinborn model and an actual scene, the monitoring and interaction real-time performance of digital twinborn is improved, and the monitoring and management requirements of a machine room can be met.

The invention relates to the field of chemical industry, and discloses a method for preparing high-purity vanadium oxide by purifying ammonium vanadate, which comprises the following steps: adding a carbonate solution into ammonium vanadate, stirring, and heating to 75-95 DEG C to obtain a vanadium-containing return solution; keeping the temperature of the vanadium-containing return solution, and adding a calcium salt impurity removal agent to obtain a vanadium-containing purified solution and first-stage impurity removal slag; oxalate is added into the vanadium-containing purified liquid, the temperature of the vanadium liquid is kept, and deeply-purified vanadium liquid and second-stage impurity removal slag are obtained; cooling the deeply purified vanadium liquid to 5-35 DEG C, stirring and reacting, and then carrying out solid-liquid separation to obtain ammonium metavanadate and secondary vanadium precipitation supernatant; adding pure water into the ammonium metavanadate product, pulping, stirring, and heating to 80-95 DEG C to obtain a secondary return solution; the pH of the secondary return solution is adjusted to 1.6-2.0, the temperature is raised to 95-98 DEG C, and ammonium polyvanadate and a third-section vanadium precipitation supernatant are obtained; and drying the ammonium polyvanadate, raising the temperature to 500-600 DEG C in an oxidizing atmosphere, and carrying out heat preservation and calcination to obtain a vanadium pentoxide product. The method can be used for preparing a high-purity vanadium pentoxide product.

The present invention provides a preparation method of 4,6-dihydroxypyrimidine, and the preparation method comprises the following steps: adding a mixed solution of malonate and formamide to an alkali metal alcoholate / solvent containing a catalyst, and then keeping the mixture warm for reaction after the addition is completed. After the reaction is completed, the solvent is evaporated to dryness, and after being dissolved in water, the mixture is acidified to a pH of 2-3, filtered, washed with water, and dried to obtain 4,6-dihydroxypyrimidine. The present invention utilizes a catalyst to catalyze the reaction, thereby improving the reaction yield, having a high atomic economy, and reducing the amount of raw materials used, reducing the generation of saline wastewater, and alleviating environmental pressure.

The invention discloses an integrated optical diversity receiving device, solves the problem of high power consumption when the influence caused by phase distortion and light beam pointing error is corrected in the existing free space optical communication, and belongs to the field of optical communication. The method comprises the steps that free space light focused by a lens is received by a multimode / few-mode optical fiber; the multi-mode light beams received by the multimode / few-mode optical fiber enter the photon lantern; the photon lantern is coupled to the single-mode fiber array after mode demultiplexing; each light beam in the single-mode fiber array enters a polarization beam splitting rotary combiner through an edge coupler; the polarization beam splitting rotation combiner performs transverse electric wave (TE) and transverse magnetic wave (TM) mode separation on the entered light beam, then rotates the separated TM polarization state light to a TE polarization state, combines the rotated TE polarization state light and the separated TE polarization state light, and then enters the MZI network; the MZI network is used for realizing optical coherent beam combination; a control circuit controls the merging of the light beams in the MZI network.

The invention discloses a supported iron-molybdenum catalyst for formaldehyde production and a preparation method and application thereof, and belongs to the field of catalysts.The catalyst comprises a core-shell structure, the core of the catalyst is a carrier, molybdenum trioxide is supported on the carrier, and the shell of the catalyst is iron molybdate. The carrier is selected from at least one of silicon dioxide, aluminum oxide, titanium oxide, magnesium oxide, magnesium silicate, a molecular sieve, white carbon black and hydroxyapatite; the preparation method comprises the following steps: firstly, loading molybdenum trioxide on a carrier, then dispersing the carrier in deionized water, dropwise adding a molybdenum source solution and an iron source solution in a parallel flow manner under stirring to perform deposition loading of iron molybdate, and performing aging, drying, forming and calcining to obtain the catalyst. The catalyst is used for preparing formaldehyde by oxidizing methanol, dimethyl ether and methylal; according to the catalyst disclosed by the invention, the raw material cost and the pollutant treatment capacity are obviously reduced due to the reduction of the dosage of molybdenum and iron components, and the catalyst has relatively good activity and stability.

The invention provides an automatic obstacle avoidance method and system based on user-defined coordinate system navigation, and relates to the technical field of automatic navigation, and the method comprises the following steps: obtaining global position information of a moving unit in a global coordinate system; judging whether the moving unit enters a new scene internal area according to the global position information; if yes, acquiring spatial morphological characteristics of an internal area of the scene into which the moving unit newly enters; according to the acquired information, judging the area type of the internal area of the newly entered scene; establishing a local coordinate system corresponding to the internal area of the newly entered scene according to the area type; and generating a local navigation path in the internal area of the scene based on the established local coordinate system. According to the method, a local coordinate system is defined, a coordinate system better matched with specific task requirements is adopted in a small range, meanwhile, the data processing amount is reduced, path planning and an obstacle avoidance algorithm are easier to implement, and higher flexibility, lower calculation cost and better precision are provided under specific conditions.

The invention provides a federal learning-based recommendation method and device, a program product and electronic equipment, and relates to the technical field of computers. The method is applied to a distributed recommendation system, the distributed recommendation system comprises a central service device and a plurality of local service devices, and the method comprises the following steps: the central service device receives respective corresponding second data sets sent by the plurality of local service devices, the second data set is obtained by performing recall processing on data in the corresponding first data set by the local service equipment by using a recall model; the central service equipment performs preliminary screening on the obtained second data set to obtain a third data set corresponding to the local service equipment, and sends the third data set to the corresponding local service equipment; and the central service equipment performs clustering sorting processing on the obtained fourth data set according to the associated information of the to-be-recommended user to obtain target information recommended to the to-be-recommended user. The recommendation efficiency and accuracy of the recommendation system are improved.

The invention provides a beneficiation and enrichment method for high-molybdenum oxide ore, which comprises the following steps of: performing fraction magnetic separation on a product after ore grinding, regrinding magnetic separation coarse particle tailings, mixing the reground product with magnetic separation fine fraction tailings to perform low-potential sulphide ore flotation, performing oxidized ore flotation on sulphide ore flotation tailings, and the like. And finally, the molybdenum concentrate with the high recovery rate and grade is obtained, the molybdenum ore separation and enrichment effect is achieved, and the fine disseminated-particle-size refractory molybdenum oxide ore resources in the high-oxidation-type molybdenum ore are effectively and comprehensively recycled. In order to quickly optimize and obtain the optimal condition of oxidized ore flotation, the flotation inhibitor with the good inhibition effect is determined through the pure ore flotation test, the flotation collecting agent with the high selectivity is selected through the molecular simulation calculation method, and the flotation efficiency and selectivity are improved. The method provided by the invention can be widely applied to the fields of development of high-oxidation molybdenum ore mines and recycling of fine-fraction refractory molybdenum oxide ore resources in tailings after sulfuration flotation.

The invention relates to a video key frame obtaining method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a plurality of candidate image frames corresponding to a to-be-processed video, carrying out the preprocessing of each candidate image frame, inputting the preprocessed candidate image frame into a pre-constructed image feature extraction model, obtaining a feature vector corresponding to each candidate image frame, and obtaining a feature vector corresponding to each candidate image frame; performing clustering processing on the feature vector corresponding to each candidate image frame through the target clustering parameter based on a preset clustering algorithm to obtain at least one target clustering result, and performing calculation based on the feature vector corresponding to each candidate image frame and the center vector corresponding to the target clustering result to obtain a target clustering result; and determining a target image frame of the to-be-processed video from the plurality of candidate image frames based on a calculation result. By adopting the technical scheme, the image frames with large scene and element differences can be screened from the continuous image frames, and the screened image frames are labeled and applied to the training of the automatic driving perception algorithm, so that the training efficiency can be greatly improved.

The invention discloses a task processing method, device and equipment running on a cloud computing platform, and relates to the technical field of cloud computing. The data of the target object can be processed by the pod side of the service container group or the intelligent hierarchical cluster, so that the situation that only one cluster or system uniformly processes a large amount of data can be avoided, and the effectiveness of data processing is improved. The method comprises the steps that firstly, a cloud computing platform obtains task information of a to-be-processed task and processing capacity information of a service container group according to task configuration operation of a user on a configuration interface; then, the cloud computing platform allocates the to-be-processed task to target computing equipment for processing based on the task information of the to-be-processed task and the processing capability information of the service container group, and the target computing equipment is the service container group or an intelligent hierarchical cluster; and finally, the intelligent hierarchical cluster generates an access data table of the target object based on a processing result of the to-be-processed task, and the access data table is used for managing a storage area of the target object.

The invention discloses a serial port data message queue processing method of gateway machine equipment and the gateway machine equipment, which adopt a Modbus protocol as a communication protocol, and store sensor data points with continuous register addresses, same read-write function codes and consistent four remote types in the same sensor into a processing data stack; the processing data heap is assembled into permanent Modbus message information; message processing instructions of the remote control information points or the remote regulation information points are assembled into temporary Modbus message information, and feedback information of the corresponding remote signaling information points and the corresponding remote measurement information points are also assembled into temporary Modbus message information; the Modbus message information is added into a Modbus message information processing queue of a serial port in a mode of combining regular traversal of a message storage pile and an upper computer instruction, and the serial port takes out the Modbus message information in the Modbus message information processing queue one by one for message processing. According to the method, the timeliness of temporary Modbus message information processing is considered while the permanent Modbus message information processing regularity is realized.

The invention relates to a mulberry disease and insect pest monitoring system and method based on Internet of Things image processing, and the method comprises the steps: obtaining mulberry disease and insect pest image data, carrying out the preprocessing and marking, creating a disease and insect pest analysis model, and training; evaluating the performance of the trained pest and disease analysis model; a first image acquisition module performs preliminary image acquisition on a to-be-monitored mulberry planting area and transmits the image to a disease and pest analysis model for preliminary disease and pest condition identification and analysis, and each disease and pest area is positioned; expanding the disease and insect pest areas, and performing infrared image acquisition on the expanded disease and insect pest areas by a second image acquisition module; identifying and analyzing pest and disease damage conditions of the expanded pest and disease damage areas; and the pest and disease drawing module draws a pest and disease region of the mulberry planting region to be monitored based on the output analysis result, renders the severity of pest and disease, and executes pest and disease control measures. Omnibearing pest and disease monitoring of mulberries is realized, and the accuracy of pest and disease recognition and analysis results is improved.

The invention belongs to the field of geological engineering construction, discloses a TBM (Tunnel Boring Machine) real-time slag monitoring system and a TBM real-time slag monitoring method for advanced geological forecast, solves the problems of lack of real-time performance, continuity and accuracy and low processing efficiency of surrounding rock state analysis in the traditional technology, and provides powerful support for advanced geological forecast in a TBM tunneling process. The method comprises the following steps: continuously acquiring rock slag images to obtain a rock slag acquisition video; cutting the rock slag acquisition video into an image sequence with a fixed time interval, and aligning the timestamp of each rock slag image with the recording time of the TBM tunneling parameters; then the cut rock slag image is preprocessed; then identifying the preprocessed rock slag image to obtain a classification result of the rock slag; and finally, the classification result of the rock slag and TBM tunneling parameters are fused to serve as input, and the surrounding rock grade is calculated through a surrounding rock grade grading model so that the tunneling strategy can be adjusted in real time. The method is suitable for underground engineering construction scenes.

The invention discloses a lightweight interest point recommendation method and device fusing user moving directions, relates to the technical field of interest point recommendation, and mainly aims to solve the problem of low data processing efficiency of interest point recommendation. The method mainly comprises the steps that in response to an interest point recommendation requirement of a target user, the real-time position, the moving direction and a global interest point of the target user are obtained, and the global interest point is a site matched with the interest point recommendation requirement in a global site sign-in record data set; screening the global points of interest according to the real-time position and the sign-in record of each point of interest to obtain a candidate sign-in record data set; calculating the moving direction similarity of each candidate interest point according to the moving direction and the real-time position, and calculating the collaborative filtering similarity of each candidate interest point according to the sign-in user information and the sign-in record of the target user; and generating interest point recommendation information of the target user according to the moving direction similarity and the collaborative filtering similarity. The method is mainly used for interest point recommendation.

The invention relates to a data processing method and device based on a large language model, equipment and a storage medium. The method comprises the following steps: acquiring input data, matching the input data with a plurality of pre-constructed data matrixes to obtain a data matrix matched with the data type of the input data, and determining the matched data matrix as a current data matrix; constructing an inference tree model based on the input data and the current data matrix; determining a plurality of candidate sequences based on the inference tree model; and performing parallel verification on the plurality of candidate sequences, determining the output probability distribution of each candidate sequence, and obtaining an output sequence matched with the data content of the input data according to the output probability distribution. Compared with the prior art, the data processing efficiency can be improved, and the data output quality and accuracy can also be improved.

The invention discloses an informatization deposit management device and method for a tourist attraction, and relates to the technical field of deposit management. After a left image and a right image are collected through a binocular camera, face area detection is carried out to obtain a left face image and a right face image; obtaining a target disparity map according to the left face image and the right face image, determining a depth change according to the target disparity map, and if the depth change is greater than a preset change, performing feature extraction on the left face image and the right face image to obtain a target feature; and determining to open the corresponding storage cabinet according to the target feature. Face image recognition is carried out on the image collected by the binocular camera to obtain the face area, then the disparity map is determined through the face image to finally obtain the depth change, whether a face recognition object is a real person or a photo can be effectively judged, the safety is improved, and finally the face recognition accuracy is improved by extracting and matching the face depth features. And the corresponding storage cabinet number is accurately found and the unlocking operation is executed, so that the access process is simplified, and the safety is also ensured.

The invention provides a power transmission project GIM model analysis loading method and device, equipment and a medium, and belongs to the technical field of power transmission engineering. A power transmission project GIM model is obtained, and tower group graph data in the power transmission project GIM model is extracted and analyzed; performing simplification processing on the tower group graphic data according to different levels, and outputting a simplification result as a three-dimensional model; slicing the three-dimensional model to form a model file; extracting state information of tower equipment in the model file; analyzing the attribute information of the tower equipment in the power transmission project GIM model; carrying out three-dimensional space arrangement on tower equipment to form a form of a power transmission line; analyzing the positions of the grounding wire and the hanging grounding wire, and storing the key point data of the grounding wire; and based on the space range control model, setting a three-dimensional model, a loading range of a grounding wire and a hanging grounding wire and a visible height of the three-dimensional model. By setting the loading range and the visible height, the loading and rendering efficiency of the model can be optimized, and the interaction experience of the user is improved.

The invention provides a campus emotion risk individual assessment method and system based on expression recognition, and the method comprises the steps: collecting monitoring image data for recognition, and obtaining recognized individual expression data and related information data; performing processing according to the individual expression data to obtain group emotion data; correcting the individual expression data according to the group emotion data to obtain individual negative emotion data; and evaluating an emotion risk target individual according to historical statistics of the individual negative emotion data. According to the method, group emotional abnormity can be accurately eliminated in a group scene, the individual emotional risk degree is objectively judged, and the accuracy of identifying the campus emotional risk target is improved.

The invention relates to the technical field of electric power data processing, in particular to an electric power data cleaning method and system based on big data. The method comprises the steps of obtaining a data fluctuation degree of each dimension through a numerical value difference between a data point in each dimension and adjacent data points on the left side and the right side; acquiring a data fluctuation degree difference value between one dimension and other dimensions to obtain a noise degree of the dimension; determining the importance degree of the dimension according to the number of data value covariance greater than 1 between the dimension and other dimensions and the difference between the dimension and other dimensions; calculating the to-be-processed possibility of each dimension; according to the method, the to-be-processed dimension is obtained through the comparison result of the to-be-processed possibility of each dimension and the preset threshold value, and the to-be-processed dimension is de-noised to realize power data cleaning, so that the power data cleaning efficiency can be effectively improved.

The invention discloses a method for refining and recovering tail gas of hydrogen chloride. The method comprises the following steps: 1) compressing the tail gas to a high-pressure state through a compressor; (2) introducing the high-pressure gas obtained in the step (1) into a refining tower for separation, obtaining non-condensable gas containing HCl at the tower top, extracting a high-purity hydrogen chloride product from the tower, and obtaining a liquid phase containing organic matters at the tower kettle; and 3) feeding the non-condensable gas at the tower top of the refining tower to a downstream absorption unit, feeding the product hydrogen chloride in the tower to the periphery for supply, decompressing the liquid phase at the tower bottom, and returning to the upstream for recycling. The method can reduce 75% of energy consumption, finally realizes purification of hydrogen chloride gas containing non-condensable gas and organic matter impurities, reduces the amount of waste gas and the amount of byproduct hydrochloric acid, produces high-purity hydrogen chloride products with high added values, realizes efficient circulation of chlorine resources in a park, and has remarkable economic benefits and social benefits.

The embodiment of the invention discloses an article recall sorting method, which comprises the following steps of: obtaining a question statement, and extracting a keyword; classifying the keywords by adopting an intelligent model to obtain field-related and irrelevant keywords; the keywords related to the field are matched with the abstracts to obtain a plurality of articles, and the articles are segmented; and performing semantic similarity sorting through vector splicing and similarity calculation. The method has the advantages that 1, the abstract generation and tagging technology is creatively applied, the database content is intelligently simplified, the data processing amount in the retrieval stage is remarkably reduced, the accuracy of preliminary screening is greatly improved, the efficiency bottleneck caused by traditional full-amount data scanning is effectively avoided, and high efficiency and accuracy of information retrieval are ensured; and 2, a multi-condition sorting algorithm is introduced, advanced technologies such as vector splicing and comparative analysis are fused, deep optimization of information extraction is realized, a most relevant result can be quickly positioned in mass data, and the quality and correlation of a final retrieval result are greatly improved.

The invention relates to the technical field of radio frequency quadrupole accelerators, in particular to a method for predicting electromagnetic field distribution of a radio frequency quadrupole accelerator by adopting a support vector machine, which comprises the following steps of: acquiring an electric field signal, a magnetic field signal and electromagnetic field distribution in the radio frequency quadrupole accelerator by presetting an electric field probe and a Hall sensor; denoising, de-trending, normalizing and smoothing the collected electric field signal and magnetic field signal to improve the signal quality and the accuracy of subsequent analysis; performing wavelet transform analysis on the collected electric field signals, extracting electric field characteristic parameters, performing wavelet transform analysis on the collected magnetic field signals, and extracting magnetic field characteristic parameters. According to the method, the wavelet transform and the support vector machine are combined, the advantages of the wavelet transform in the aspects of multi-scale signal analysis and noise suppression can be fully played, and meanwhile, the precision, efficiency, robustness and adaptability of radio frequency quadrupole accelerator electromagnetic field distribution prediction are remarkably improved by utilizing the strong classification and regression capability of the SVM.