127results about How to "Improve guidance" patented technology

The invention discloses a distributed large model training-oriented multi-granularity semantic tree construction system and a distributed large model training-oriented multi-granularity semantic tree construction method. According to an edge-cloud information fusion space, samples and text tags of a cloud server are embedded and collected into public data set information of edge equipment # imgabs0 # for multi-modal data fusion; the model information interaction space is used for embedding samples and text labels of a cloud server and obtaining a global initial cluster center Ol belonging to Rd * l by adopting a hypergraph clustering algorithm; the cloud server constructs a semantic space # imgabs1 # based on the initial cluster center Ol, and the cloud server distributes the initial cluster center Ol and a text embedding # imgabs2 # of the semantic space to each edge device to construct an edge-cloud cooperative interaction space; the cloud server generates a multi-granularity semantic tree; according to the edge-cloud collaborative interaction space construction global granularity generation model, appropriate data granularity is cooperatively generated in a construction system, and the matching requirement of a heterogeneous model is met.

The invention relates to the technical field of data processing, in particular to a retrieval question and answer method applied to a large language model. In the embodiment of the invention, the electronic equipment retrieves a target information block similar to a question text based on a TF-IDF retrieval method, retrieves a target document fragment similar to the question text in the target information block based on a keyword and a vector, and takes the target document fragment as a knowledge text used for replying the question text, so that the question text is replied to the question text. According to the method, the retrieval effectiveness is improved, the influence of redundant knowledge is reduced, the cue words are dynamically constructed based on the retrieved knowledge text, and the question and answer retrieval accuracy is improved.

The invention provides a fire coal batch management method, device and equipment based on point cloud data, and the method comprises the steps: scanning a coal yard through a laser coal checking instrument to obtain coal pile surface information data, converting the coal pile surface information data into spatial three-dimensional coordinate data, stripping various coal in the coal yard, and carrying out the structure presentation of the coal pile through a three-dimensional structure; acquiring coal piling and taking scheduling task information, coal piling and taking batch information and coal quality composition information of the fuel management system and the generation management system in the same time period, and determining an acting coal area in a scheduling task; comparing the acting coal area in the scheduling task with each coal area in the coal pile structure, if the coal area information is the same, determining that the coal area in the corresponding coal pile structure is the coal quality in the scheduling task, and completing coal quality information matching and binding; based on matching and binding, real-time display of three-dimensional models of different batches of coal piling and coal taking in a coal yard is realized, so that management of fire coal is realized more accurately, and a good guiding effect on coal blending is also facilitated.

The embodiment of the present invention discloses a data processing method, device, equipment and storage medium based on sentiment analysis, the method comprising: obtaining financial field review text data of a target date; inputting the financial field review text data into a hybrid neural network to obtain sentiment analysis results of the financial field review text data of the target date; generating financial market trend data of a target associated date according to the sentiment analysis results of the financial field review text data of the target date. The technical solution of the embodiment of the present invention can realize the automatic analysis and processing of financial field sentiment data, improve the accuracy and efficiency of the analysis and processing of financial field sentiment data, and thus improve the guidance and applicability of financial field sentiment data.

The invention discloses a shoreline classification method based on erosion and deposition and form quantitative analysis, and the method comprises the steps: carrying out the longitudinal discretization of a shoreline, and determining the transverse topographic data of each discrete point on the shoreline; calculating erosion and deposition characteristic parameters and morphological characteristic parameters of each discrete point based on the transverse topographic data of each discrete point; drawing a shoreline erosion and deposition characteristic distribution diagram and a shoreline morphological characteristic distribution diagram based on the erosion and deposition characteristic parameters and the morphological characteristic parameters; and determining the type of each section on the shoreline according to the shoreline erosion and deposition characteristic distribution diagram, the shoreline morphological characteristic distribution diagram and the shoreline characteristic matrix. According to the shoreline classification method, the recent stable state of the shoreline can be visually expressed on the basis of erosion and deposition quantification and form quantification analysis and calculation of the river bank slope, and a good guiding effect is achieved on shoreline development and utilization behavior selection, flood control project safety evaluation, shoreline bearing capacity evaluation, shoreline protection improvement optimization and the like.

The invention discloses a casting-infiltration depth prediction method for preparing a metal-based composite material based on a casting-infiltration method, and relates to the technical field of material forming. The casting-infiltration depth prediction method comprises the steps that the casting-infiltration initial temperature at which molten metal starts to infiltrate into a porous prefabricated body and the casting-infiltration ending temperature at which the molten metal stops infiltrating into the porous prefabricated body are determined; calculating the heat loss of the molten metal in the process of infiltrating the porous preform according to the casting infiltration initial temperature and the casting infiltration finishing temperature; the casting infiltration duration of the molten metal is calculated according to the heat loss of the molten metal; and calculating the theoretical casting infiltration depth of the molten metal in the porous preform according to the casting infiltration duration and the casting infiltration speed of the molten metal. According to the casting-infiltration depth prediction method provided by the embodiment of the invention, the theoretical casting-infiltration depth of the metal-based composite material prepared by the casting-infiltration method can be effectively predicted, and the obtained theoretical casting-infiltration depth has a relatively good evaluation effect and a relatively good guidance effect on the casting-infiltration process.

The invention discloses a tea production cycle monitoring technology based on remote sensing images, particularly relates to the field of remote sensing image calculation, and comprises a monitoring area division module, a monitoring data acquisition module, a monitoring data analysis module, a comprehensive analysis module and an early warning module. According to the invention, through a tea production period monitoring technology based on remote sensing images, accurate monitoring and management of the whole production process of the tea garden are realized, through monitoring area division, data acquisition, analysis, comprehensive evaluation, early warning and other modules, the accuracy and real-time performance of monitoring data are improved, the scientificity and efficiency of tea garden management are enhanced, and in addition, the tea production period monitoring system is suitable for popularization and application. Potential problems can be found and early warned in time, powerful decision support is provided for tea garden managers, the quality and benefits of tea production are improved, and sustainable development of the tea industry is promoted.

The invention relates to the technical field of data processing, in particular to a map data fusion method, a vehicle, equipment and a program product. The method comprises the steps that lightweight map data and high-precision map data of a target area are acquired, and the number of road elements contained in the high-precision map data is larger than the number of road elements contained in the lightweight map data; extracting target high-precision road elements used for being fused to lightweight map data from the high-precision map data; carrying out adaptive processing on the target high-precision road element to obtain a target lightweight road element adapted to the lightweight map data; and fusing the target lightweight road element into the lightweight map data to obtain fused map data. According to the method, the lightweight map data and the high-precision map data are integrated, and the required target lightweight road elements are fused in the lightweight map data, so that the finally obtained fused map data can be rapidly loaded or updated, and the road elements in the map can be ensured to be more comprehensive.

The utility model belongs to the technical field of flue gas testing, and particularly relates to a container type flue gas testing device. According to the technology, a container is included, the container is internally divided into an equipment area, a center control and office area and an operation area, an equipment installation table is installed in the equipment area, and an operation table, an experiment device area and an instrument area are arranged on the equipment installation table; wherein the experimental device area and the instrument area are mounted on the equipment mounting table on the rear side of the operation table and are distributed left and right. The whole design is of a container structure, the structure is compact, attractive and elegant, the size is small, transportation is convenient, tests can be carried out in different regions and under different flue gas conditions according to actual needs, the utilization rate of the device is greatly improved, and rebuilding and reinvestment are avoided; meanwhile, the same device is used for testing under different flue gas conditions, and only flue gas is used as a variable, so that the test result difference caused by the test device difference is avoided.

The invention relates to an integrated bionic double-layer anti-bending artificial blood vessel and a preparation method thereof.The artificial blood vessel is of a hollow tubular fabric structure and is composed of an inner layer and an outer layer, and the inner surface of the inner layer is a smooth curved surface formed by interweaving multifilaments; the outer layer is of a fluffy, porous and circumferentially oriented arrangement structure formed by hooking, stretching and deforming monofilaments in the multifilaments on the outer surface of the inner layer, and the inner layer and the outer layer are of a tightly connected integrated structure; during preparation, multifilaments are adopted to form the tubular skeleton of the artificial blood vessel through a weaving, knitting or weaving method, napping, snagging or pile cutting is firstly carried out after treatment, then a carding process is carried out, and finally post-treatment shaping is carried out, so that a product is obtained. The preparation method is simple, and the double-layer artificial blood vessel is integrally formed and has good bending resistance; the inner layer and the outer layer of the artificial blood vessel are of an integrated structure, multi-layer structure regeneration of the blood vessel is promoted through the design of the double-layer structure artificial blood vessel, the problem of joint surface layering does not exist during degradation, and a stable structure for cell adhesion growth in the blood vessel regeneration process is provided.

The invention discloses an English teaching writing error diagnosis method based on artificial intelligence, and relates to the technical field of artificial intelligence, and the method comprises the following steps: obtaining an English composition text, and obtaining initial writing data; performing grammatical structure analysis according to the initial writing data to obtain grammatical features; performing vocabulary error analysis according to the grammatical features to obtain vocabulary errors; performing syntactic and semantic consistency analysis according to the vocabulary errors to obtain syntactic and semantic errors; performing article logic analysis according to the syntactic and semantic errors to obtain article logic errors; and performing comprehensive diagnosis according to the article logic error to obtain a diagnosis result. According to the method, the conjunction matching degree and the semantic compatibility are combined, cross-sentence logic analysis is achieved, misjudgment of isolated sentences is avoided, a complete error set is formed, the comprehensiveness and accuracy of diagnosis are improved, and practicability and guidance are enhanced.

The invention discloses an intelligent supervision method and system for a building engineering project, a medium and a product, and the method comprises the steps: determining expected time-workload-material progress according to workload information, material information and a construction plan of the engineering project; the actual construction time progress and the actual construction workload progress sent by the construction site end are compared with the preset construction time progress and the preset construction workload progress, and whether the actual construction time progress and the actual construction workload progress are within the range of the preset construction time progress and the preset construction workload progress or not is judged; if yes, the actual material use progress is determined according to the RFID tag, sent by the construction site end, of the used building material, the actual material use progress is compared with the preset construction material progress, and whether the actual material use progress is within the preset construction material progress range or not is determined; and if not, building material use early warning information is sent to the user management end. By implementing the technical scheme provided by the invention, the supervision efficiency and accuracy can be improved.

The invention provides a code coverage rate obtaining method and device and electronic equipment, and the method comprises the steps: obtaining a test code file and a code coverage rate statistical scene to which the test code file belongs; determining a target code snippet participating in code coverage rate statistics in the test code file based on the code coverage rate statistics scene; running the test code file to obtain a running code snippet of the test code file running on the test terminal; and obtaining a target code coverage rate of the test code file based on the target code snippet and the operation code snippet. According to the method, the abnormal condition that the redundant code snippets in the test code file are included in the code coverage rate statistical range is avoided, the code coverage rate obtaining method is optimized, the accuracy degree of the code coverage rate is improved, and the guidance and reference of the code coverage rate are optimized.

The invention discloses an intelligent data processing method for ship berthing, and the method comprises the steps: respectively detecting and judging the wind speed values of a bow end, a stern end and a middle section when determining that the wind directions of the bow end and the stern end of a ship are the downwind directions facing an embankment; the safety determination value is determined based on different judgment results, and the output safety signal is determined based on the magnitude relationship between the safety determination value and the safety judgment value; further, the downwind at different positions in the length direction of the ship body is analyzed and judged, and safety prompt feedback is performed on a driver in the berthing process through intelligent analysis and processing, so that intelligent data analysis and processing during berthing of the ship are realized; the safety and guidance in the ship berthing process are improved.

The utility model provides a multi-position synchronous pressing tool which comprises a bottom plate, a positioning block, a left pressing block, a right pressing block, a push-pull block and a pull rod. In the using process, a pressed workpiece is inserted into the positioning block from front to back till the workpiece cannot move, workpiece positioning is achieved, then the pull rod is rotated clockwise, the push-pull block is pushed to move leftwards through threads of the connecting block, and in the process that the push-pull block moves leftwards, the left pressing block and the right pressing block are pressed downwards at the same time under the matching effect of the inclined faces; in the process that the left pressing block and the right pressing block are pressed downwards at the same time, the inclined slope of the push-pull block and the inclined tables of the left pressing block and the right pressing block extrude each other under the action of thrust, and therefore pressing and positioning of the workpiece to be pressed are achieved. During production changing, the pull rod is rotated anticlockwise, the pull rod moves rightwards through the threads of the connecting block, the push-pull block is pulled to move rightwards, and in the process that the push-pull block moves rightwards, the left pressing block and the right pressing block move upwards, and a workpiece is loosened. According to the synchronous pressing tool, only simple inclined surfaces of the left pressing block, the right pressing block and the push-pull block are matched and extruded, and the push-pull block, the pull rod and the connecting block are directly of a threaded structure to form a thrust effect, so that synchronous pressing of multiple positions of the tool is realized, the workpiece clamping efficiency and the product quality are greatly improved, the use cost is effectively reduced, and the production efficiency is improved. And the method has a good guiding effect on other similar working conditions.

The invention discloses a noise model circuit based on a time-interleaved pipeline ADC system, which relates to the technical field of integrated circuits and comprises an isolation amplifier noise model, a reference circuit noise model and a time-interleaved pipeline ADC model; the isolation amplifier noise model is composed of an operational amplifier and an external feedback structure, and processes the thermal noise of the operational amplifier and its feedback network; the reference circuit noise model processes the thermal noise and quantization noise generated by the reference circuit; the time-interleaved pipeline ADC model comprises a time-interleaved circuit and a pipeline ADC, and processes the noise generated by the sampling structure and the operational amplifier; the invention not only includes a traditional ADC structure in structure, but also innovatively considers the influence of the isolation amplifier and the power supply, which has a better guiding role in building a highly integrated ADC system.

The utility model discloses a transformer hoisting device mainly relates to the power station equipment installation technical field. Including the beam, the both ends of beam top side are equipped with the upper lifting hole respectively, the upper lifting hole is hung with the lifting cable, the both ends of beam bottom side are equipped with the lower lifting hole respectively, the lower lifting hole is hung with the lifting belt or the lifting cable, the both ends of beam are equipped with the length direction same direction extension's rebates respectively, the rebates are hinged with the swing arm, the inner end of swing arm is located in the rebates and the outer end of swing arm exceeds the both ends of beam, the outer end rotatable mounting of swing arm has the vertical guide pulley of axial, be equipped with the elastic element between swing arm and rebates, the elastic element makes swing arm keep with beam and keep static same direction, and when swing arm swings, the elastic element has the elastic force of making swing arm reset for with beam same direction. The utility model has the beneficial effect at in narrow space to the hoisting path can be in the middle direction, prevents and twists and turns and hangs the partiality.

The present application relates to the technical field of intelligent construction control, in particular to a beam segment hoisting assembly precision detection method and system based on improved machine vision, comprising the following steps: a visual guidance zone extending along the joint is formed in advance at the assembly interface of the to-be-hoisted beam segment and the butt joint beam segment, and a joint mark with a direction recognition relationship is arranged in the visual guidance zone; during the process of the to-be-hoisted beam segment approaching the butt joint beam segment, the visual guidance zone is continuously imaged, and according to the changes of the joint mark in the continuous images, the changes of the joint mark in the continuous images, and the changes of the relative pointing direction, the joint evolution result of the corresponding assembly interface is generated. The present application realizes explicit mapping of the gap, height difference and lateral misalignment, breaks through the instability of traditional machine vision relying on natural edge recognition, and improves the accuracy and robustness of joint state recognition in complex construction environment.

The invention discloses a new energy station operation evaluation model equivalent modeling method, and belongs to the field of power systems. The problem that in the existing equivalent modeling process of the new energy station, the identifiability of data is too low is solved. The method comprises the steps of extracting initial variable indexes of a new energy unit; establishing a factor analysis model of the initial variable indexes; obtaining factor score data as a grouping index; performing similarity measurement on the grouping indexes by using correlation analysis and significance test, and obtaining a correlation relationship of the new energy units to complete group division; and calculating equivalent parameters based on a grouping result, and establishing a new energy station equivalent model. The method is mainly used for carrying out equivalent modeling on the new energy station.

The invention discloses a drilling process state monitoring method based on interval information granulation, and relates to the technical field of geological drilling engineering.The drilling process state monitoring method based on interval information granulation mainly comprises the steps that original drilling process data is preprocessed to obtain a drilling data time sequence; performing multi-interval information granulation on the drilling data time sequence to obtain a low-dimension time sequence; according to the low-dimension time sequence, a drilling operation state monitoring model is obtained through a principal component analysis method with a kernel function; and monitoring the drilling operation state by using the drilling operation state monitoring model. According to the drilling process state monitoring method based on interval information granulation, the operation state of the drilling system can be finely monitored and sensitively sensed.

The application relates to a strip steel surface defect tracing method, system, device and medium. The method comprises the following steps: performing defect recognition on a strip steel surface defect image to obtain a defect type, a geometric feature vector and a texture feature vector, acquiring spatial coordinates and a timestamp corresponding to the image, and constructing a defect feature matrix; constructing a process knowledge graph based on production line equipment parameters, a theoretical speed curve, equipment topological relations, real-time working condition data and historical event records; performing candidate equipment positioning in the process knowledge graph by adopting a periodic matching strategy or a non-periodic matching strategy to generate a matching result set; acquiring an equipment health index from the process knowledge graph, calculating a final confidence and a comprehensive score based on matching parameters in the matching result set, and generating a tracing report. The method integrates multi-dimensional defect features and constructs a process knowledge graph, optimizes a defect and equipment matching strategy, and improves the accuracy and reliability of strip steel surface defect tracing.

The present application relates to a semantic chain feedback and reinforcement learning driven image fusion method, belonging to the technical field of computer vision and image processing. It includes: inputting the preprocessed input image into the frozen dual-path hierarchical encoder, extracting multi-scale visual features and shallow prior features in the prior injection branch; on each intermediate scale, cross-modal collaborative fusion is performed on the infrared features, visible light features, the last layer propagation features and shallow prior features in the multi-scale visual features; chain semantic enhancement is performed on the multi-scale fusion features: obtaining several candidate fusion images; optimizing the unfrozen model parameters; inputting the infrared image and visible light image to be processed into the optimized network, and selecting the noise-free reference candidate as the final fusion image output. This method cooperatively optimizes through chain semantic progressive modulation and difference driven reinforcement learning, improving the visual quality and information expression ability of the fusion image.

The present invention provides a maintenance plan adjustment method and system based on constants and variables, the method comprising: determining constants and decision variables; establishing a rail transit linear asset grid maintenance plan optimization adjustment model with a plan change minimum objective function according to the constants and decision variables; adjusting the maintenance plan according to the rail transit linear asset grid maintenance plan optimization adjustment model. The present invention enhances the guidance and practicality of the rail transit linear asset maintenance plan.

The invention provides a mutual inductor operation and maintenance method based on an improved K-means algorithm, and the method comprises the steps: collecting the operation time and an operation time ratio difference sequence of a mutual inductor, calculating the reliability of the mutual inductor through three-parameter Weibull distribution, and evaluating the stability of the mutual inductor through ADF statistics; the reliability and the stability are used as key indexes, a mutual inductor data set is constructed, kernel density estimation (KDE) and a local outlier factor (LOF) are adopted to optimize initial cluster center selection of a K-means algorithm, and particularly, by calculating a cluster center variation coefficient and a local reachable density, it is ensured that the dispersity and the density of cluster center distribution are reasonable. After a mutual inductor data set is subjected to clustering analysis through an improved K-means algorithm, the reliability and stability mean value of each cluster is calculated, weights are given according to user requirements, comprehensive scores are calculated, and the comprehensive scores are divided into five grades to evaluate the state of the mutual inductor. And finally, different operation and maintenance strategies are made according to the evaluation result, and the accuracy and reliability of the operation and maintenance of the mutual inductor are improved.

The invention discloses a double-variable-slope internal circulation water tank simulation experiment device and method, and relates to the technical field of deposition simulation testing. The device comprises a water tank, at least two stages of slope changing devices are arranged in the water tank, and a water inlet of the water tank is higher than a water outlet of the water tank; the injection system comprises a slurry pump, the inlet end of the slurry pump is connected with a slurry pool, and the outlet end is connected with the water inlet of the water tank; the circulating system comprises an overflow pool, a collection hopper is arranged at the bottom of the overflow pool and connected with the water inlet of the water tank through a circulating slurry pump, and one end of the overflow pool is communicated with the water inlet. And during an experiment, internal circulation can be carried out through the circulating system, so that the cost is saved. According to the device and the method disclosed by the invention, a plurality of variable slopes are arranged, so that a gravity flow deposition experiment under a multi-slope folding terrain condition can be simulated; and meanwhile, a circulating system is arranged, so that rapid deposition of a large amount of silt caused by long-time injection of mud into the single-flow water tank is prevented, environmental protection is achieved, and raw materials are saved.

The application discloses a kind of engineering cost auditing method and system based on multi-source data fusion.Belongs to the field of engineering cost auditing, the present application obtains the engineering quantity list of the engineering to be audited, construction site space-time image and construction log text data, input model trained by multi-modal data, generates cost declaration and physical live characteristic vector, then calculates correlation score and extracts empirical segment features, and with cost declaration features fusion generates semantic fusion feature matrix;Finally, identify sub-item engineering category, calculate its Mahalanobis distance with process feature prototype vector, if exceed threshold value, then determine abnormal, generate the audit result containing abnormal cost item and construction period lack of empirical information.The present application effectively solves the problem that existing technology cannot accurately and efficiently audit engineering cost.

The invention discloses a meat freshness rapid detection device, and belongs to the technical field of food nondestructive detection. The device aims at solving the problems that existing detection equipment is poor in portability and difficult to use on site, the detection range is limited due to single excitation wavelength, and the detection process depends on manual operation and is insufficient in intelligence. According to the technical scheme, the device is characterized by comprising a sample placement area, an annular multi-wavelength excitation light source module, a spectrum acquisition module, a single-chip microcomputer module, a Bluetooth module and a power supply and discharge module which are integrated in a device shell; the external intelligent terminal is provided with detection analysis software, and the software controls the excitation of the LED lamp beads with specific wavelengths through the Bluetooth module, automatically identifies the type of the freshness indication label based on the collected spectral characteristics, and calls a corresponding prediction model to calculate the freshness index. The device is mainly used for carrying out rapid and portable on-site detection on meat samples with freshness indication labels.

The embodiment of the invention relates to the field of coal-fired power generation, and provides a boiler convection heating surface cleaning factor calculation and test system and method, and the system comprises the steps: employing a configuration function block of a production control system through employing the existing thermotechnical measurement point of a coal-fired power generation unit, and according to the design parameters and historical operation characteristic parameters of the coal-fired power generation unit, calculating a boiler convection heating surface cleaning factor; a real-time calculation loop is embedded in a production control system, working conditions are automatically screened when the coal-fired power generation unit continuously operates, typical test working condition signals of the coal-fired power generation unit and long-acting cleaning factor signals of a certain convection heating surface of a boiler are output, and the purpose that in the fully-closed production control system, the real-time calculation loop can be automatically calculated is achieved. On-line monitoring of the boiler convection heating surface cleaning factor can be realized without third-party auxiliary software, a good guiding effect on safe and stable operation and adjustment control of the boiler convection heating surface is achieved, and the investment cost of thermotechnical measuring points, third-party monitoring and other software and hardware is greatly reduced. And autonomous optimization and maintenance can be carried out in time according to working condition changes, so that the maintenance cost is reduced.

The present invention discloses a method for comprehensive evaluation of the compressibility of shale oil horizontal well volume fracturing, which is specifically as follows: step 1, using the analytic hierarchy process to establish a multi-level evaluation system of factors affecting the compressibility of shale oil reservoirs; step 2, according to the logging interpretation results of the horizontal section of the horizontal well, the geological parameters and rock mechanics parameter sets described in step 1 are obtained, and the maximum and minimum values ​​of each parameter in the geological parameter set and the rock mechanics parameter set are determined; step 3, the threshold of the reservoir section with the best compressibility is determined, and the best compressibility reservoir section is established; step 4, a shale oil reservoir compressibility evaluation model is established, and the comprehensive evaluation factors of different reservoir sections are calculated to quantitatively evaluate the compressibility of the reservoir. The evaluation method of the present invention considers both geological factors and rock mechanics factors to evaluate the compressibility of shale reservoirs, and realizes the quantitative evaluation of the compressibility of shale oil horizontal well volume fracturing.