115results about How to "Improve processing precision" patented technology

The invention discloses a method for correcting dialogue intention information of old people based on a deep neural network, and particularly relates to the technical field of dialogue optimization, and the method comprises the steps: collecting and cleaning multi-topic multi-round dialogue data, marking multi-round dialogue corpora, and unifying the format; performing feature extraction on the labeled corpus by using the multi-intention recognition sub-model to generate a multi-intention representation vector capable of representing each intention; screening context content related to the target intention in combination with a hierarchical attention mechanism of user context dynamic weight adjustment; processing context content related to the target intention through gating operation based on multiple rounds of interaction history, and reserving semantic fragments with reference value; performing parallel analysis on the reference semantic fragments based on a multi-task learning structure of a deep neural network, performing classification processing on different intentions, and calculating scores; and executing multi-intention fusion correction operation according to classification and scoring results, and outputting intention information capable of distinguishing each intention.

The invention discloses a medical image data processing method and system, and relates to medical image data processing, and the method comprises the steps: introducing an attention mechanism to calculate the attention weight of each mode based on the specific features of each mode, generating a specific feature map according to the attention weight, and carrying out the recognition of the specific feature map through a feature alignment algorithm, performing spatial alignment on the specific feature maps of the different modalities to generate an aligned specific feature map, and identifying correlation among the modalities by using a multi-head attention mechanism to generate a comprehensive feature map; and based on the comprehensive feature map, carrying out pixel classification through UNet to generate a preliminary segmentation result of the lesion area, meanwhile, utilizing an edge monitoring algorithm to obtain a multi-scale edge feature map of the lesion area, comparing the multi-scale edge feature map with the preliminary segmentation result, removing a pseudo lesion area according to a comparison result, and finally generating a lesion positioning map. According to the method, the pseudo focus area is eliminated through multi-scale edge feature comparison, and the accuracy and robustness of focus positioning are ensured.

The invention discloses an intelligent interactive Agent system based on Langchain and an implementation method, and relates to the technical field of natural language processing and artificial intelligence. Comprising the steps of 1, establishing an intelligent interactive Agent system by using an interface and an integrated tool provided by a Langchain framework, 2, managing calling of various tool modules and services through a central scheduling module, and dynamically registering and unbinding the tool modules by using the interface provided by the Langchain framework, and 3, processing user input by using a large language model through a process management module, and establishing an intelligent interactive Agent system. 4, a series of tool modules are integrated through the integrated tool management module, the tool modules comprise a text search tool, a weather query tool and a video information acquisition tool, each tool module works independently or is in linkage with other tool modules, and the tool modules are connected with the integrated tool management module. And 5, managing a user interaction interface through a front-end management module, supporting a user to query and control in a natural language, displaying a processing progress and a processing result in real time, and providing feedback and help information.

The invention relates to the technical field of television image communication, in particular to a user information input method and system in an intelligent television all-in-one machine. The method comprises the following steps: collecting a user image, carrying out white balance correction and focusing processing, generating a moving object focusing image, carrying out identification and attention tendency analysis on a character in the focusing image, obtaining character tendency data, analyzing a historical interaction image, identifying an interaction mode and constructing a user intention prediction model, and obtaining a user intention prediction result. And when the tendency data shows that the user preferences the television content, intention drilling and auxiliary adjustment are performed on the current image based on the prediction model, predicted user input information is generated, logic conflicts in the predicted information are detected, feedback optimization is performed, and final user input information is output. According to the invention, a more accurate and more flexible user information input method is realized.

The embodiment of the invention discloses a new infectious disease monitoring and early warning method and device based on multi-point triggering. The method comprises the following steps: acquiring clinical data of a patient in HIS, LIS, PACS, operation and electronic medical records from a medical institution; setting a screening strategy of new infectious diseases; identifying suspected cases according to the clinical data of the patient; identifying abnormal diagnosis and treatment behaviors of the patient from clinical data of the patient based on big data analysis; determining a suspected case level according to the suspected case and the abnormal diagnosis and treatment behavior; obtaining relevant survey data of suspected cases; judging whether the suspected cases have aggregated cases or not according to the related survey data of the suspected cases; and if the suspected cases have aggregated cases, determining an infectious disease transmission route, and performing case management. By implementing the method provided by the embodiment of the invention, real-time monitoring and early warning of infectious disease epidemic situations and other public health events can be realized, and the processing efficiency and accuracy of new infectious diseases are improved.

The invention relates to the field of image processing, and discloses a violation image detection method and device, equipment and a storage medium. The violation image detection method comprises the following steps: acquiring pre-annotation data obtained by annotating a sample image based on a preset model auxiliary annotation system; training an image processing model of a preset decoupling framework based on the pre-annotated data to construct a violation image detection model; and performing violation detection on a target image based on the violation image detection model, the target image being obtained by preprocessing an original image based on a preset redundancy removal algorithm. By using the preset model auxiliary labeling system, the preset decoupling framework and the preset redundancy removal algorithm, the positioning and classification efficiency of labeling and the model training efficiency can be remarkably improved, the model deployment cost is reduced, the model processing efficiency is improved, efficient violation detection of the image is realized, and the detection accuracy is improved. The risk control scene of the vertical screen picture of the mobile phone can be effectively compatible.

The invention relates to a red-black image recognition and extraction method and system based on a power distribution network line, and belongs to the technical field of data processing. The method comprises the following steps: acquiring a red image and a black image of a power distribution network line from a GIS system; traversing the red image and the black image based on a set traversal algorithm, identifying switch equipment nodes in the power distribution network, and constructing a topological structure of the power distribution network; wherein a traversal algorithm is set to express the power distribution network as a graph, then node traversal is carried out, and verification is carried out through a preset rule; and based on the power distribution network topological structure, automatically constructing a power distribution network sketch. According to the embodiment of the invention, a complex and disordered power distribution network red-black diagram is converted into a power distribution network sketch which is matched with the switch equipment nodes more conveniently and visually, a solid foundation can be laid for subsequent intelligent monitoring of the power distribution network, and the data processing precision is improved.

The invention discloses a welding quality on-line detection method based on a laser welding machine, and relates to the technical field of welding process quality detection.The welding quality on-line detection method comprises the steps that a welding seam image in the welding process of the laser welding machine is collected, the welding seam image is converted into a gray level image, and binarization processing is conducted on the gray level image; performing morphological processing on the binarized image, eliminating noisy points in the image through opening operation and closing operation, and extracting edge contour features of the weld seam; detecting a straight line segment in the weld edge contour features, and calculating geometric parameters of the straight line segment; and analyzing the geometric parameters by adopting a quality scoring model to generate a weld quality score value. Through improved image processing, an efficient and accurate solution is provided for weld joint quality evaluation in the laser welding process, the accuracy of weld joint quality detection is improved, the real-time performance and the intelligent level of the evaluation process are optimized, and the method has wide application prospects.

The invention relates to the technical field of power plant inspection, in particular to an information processing method and device of an inspection robot, equipment and a storage medium, and the method comprises the steps: collecting an equipment image in a moving inspection process of the inspection robot, carrying out the instrument detection based on the collected equipment image, and determining an instrument region image in the collected equipment image; the method comprises the following steps: reducing noise influence in the environment, correcting and identifying an instrument area image, extracting instrument information of each equipment image acquired in the inspection process, improving the information processing precision of an inspection robot in the inspection process, and finally performing equipment self-inspection according to the equipment instrument information, thereby improving the safe operation capability of equipment in a power plant. The technical problem that in the prior art, the inspection efficiency of equipment in a power plant is low is solved.

The invention relates to a data processing method and device based on a hybrid expert model, computer equipment, a computer readable storage medium and a computer program product. The method comprises: based on a target character in input data, determining a target probability corresponding to the target character, and determining a target probability difference threshold corresponding to the target character, the target probability comprising a probability estimation value of each expert network model in a preset hybrid expert model; screening the expert network models based on the probability estimation values of the expert network models and a target probability difference threshold to obtain a target expert network model; and processing the target character based on each target expert network model to obtain a data processing result. By adopting the method, the models contained in the expert model can be dynamically mixed based on the probability difference threshold, the data processing of the character dimension of the input data is realized, and the adaptation degree between each dynamically determined expert network model and the characters in the input data is ensured.

The invention relates to a semi-supervised underwater sound event detection method based on joint disturbance consistency constraint. The method comprises the following steps: acquiring logarithmic Mel spectrums of strong label audio data, weak label audio data and unlabeled audio data; performing weak enhancement and strong enhancement processing on the logarithmic Mel spectrum of the unlabeled audio data to obtain a corresponding enhanced spectrum; constructing a deep learning network model, and performing supervised training by using the frequency spectrums of the strong label audio data and the weak label audio data; adopting a multi-feature disturbance and strong disturbance combined semi-supervised model for the unlabeled audio data; and finally, optimizing the training process of the deep learning network model by calculating a joint loss function. And detecting an underwater sound event by using the trained deep learning network model, and outputting an event category of the audio data. The method is used for realizing effective detection of the sound event by using a small amount of strong injection data and improving the model performance.

The invention discloses a bridge point cloud extraction method based on evolution gradient search, and relates to the technical field of data processing, and the method comprises the steps: employing a clustering algorithm to carry out the denoising of a bridge surface point cloud; preliminarily extracting the de-noised bridge floor point cloud based on a multi-thread and minimum heap data structure, and meanwhile, setting a perpendicularity density constraint to further extract the bridge floor point cloud; a gradient-based optimization algorithm is combined with an evolutionary strategy to serve as a complementary algorithm to be incorporated into a unified framework, and the extracted bridge deck point cloud is optimized by alternately adopting a gradient updating method and an evolutionary updating method, so that the finally extracted bridge deck point cloud is obtained. According to the method, the processing precision of the bridge deck point cloud is improved, the main body structure of the bridge can be more effectively identified and separated from the complex point cloud data by introducing the evolution gradient search strategy, and meanwhile, the interference of noise and redundant information on a final result is remarkably reduced.

The invention relates to the technical field of intelligent mine mining, in particular to an intelligent mine physical property parameter real-time inversion and coal mining simulation real-time optimization method. The method specifically comprises the steps of arranging monitoring points, and monitoring physical property parameter stress data sigma and displacement data u in real time; establishing a mathematical model composed of a geometric equation, an equilibrium equation, a constitutive equation and initial-boundary conditions; a solver is used for simulating the coal mining process, and simulated stress data sigma simulation and displacement data u simulation of each monitoring point are obtained; comparing the real-time simulation stress data sigma simulation and the displacement data u simulation of each monitoring point with the field real-time monitoring stress data sigma monitoring and the displacement data u monitoring, and constructing a loss function; and by utilizing an optimization algorithm, the physical property parameters of the stratum are determined while the coal mining simulation is more accurate. According to the method, the accuracy and optimization effect of coal mining simulation can be improved, so that the safety and resource utilization rate of mine mining are improved.

The invention relates to the technical field of signal processing, and discloses a harbor equipment vibration signal processing method, which comprises the following steps: constructing a comprehensive function, and calculating a harbor equipment vibration signal based on the characteristics of a frequency scale, a second wavelet basis function and a threshold value; the concentration ratio of energy distribution, the calculation complexity of second wavelet basis functions under the frequency scale, the orthogonal value between the second wavelet basis functions under the frequency scale, the sparse value of second wavelet coefficients under the frequency scale and the signal-to-noise ratio of time domain signals reconstructed after threshold processing are comprehensively considered. The selection of the optimized frequency scale, the optimized second wavelet basis function and the optimized threshold is realized. The innovative method gets rid of the limitation of a single or optimized algorithm, fixed parameter setting or predefined rules in the prior art, and significantly improves the precision and robustness of low-frequency vibration signal processing.

The invention provides a measuring point selection method and system based on a 3D radar level gauge, and relates to the technical field of measurement, and the method comprises the steps: obtaining original 3D radar point cloud data of a measurement scene, carrying out the adaptive space division of the measurement scene through employing an improved octree algorithm, obtaining a three-dimensional reconstruction result of the measurement scene, and carrying out the calculation of the three-dimensional reconstruction result. Performing normal vector calculation and region clustering, and determining a plurality of key regions in the three-dimensional reconstruction result in combination with a region growing and merging algorithm; and taking maximization of the coverage degree of the measurement points to the key area as a first target, taking minimization of the geometric distribution difference degree between the measurement points as a second target, solving the single-target optimization problem through an improved differential evolution algorithm, and selecting a solution with an optimal comprehensive evaluation index as a final 3D radar level gauge measurement point optimization arrangement scheme.

The invention discloses a homeland space planning data intelligent analysis method based on big data, and relates to the technical field of homeland space planning, and the method comprises the steps: collecting multi-source data, carrying out the quality screening, and forming a standardized data set. And setting a space element extraction rule based on a specific planning target, and mining key space elements from the standardized data. Constructing a big data analysis model, inputting key elements to calculate and obtain a spatial development tendency index, and generating a preliminary planning scheme according to the spatial development tendency index; and performing compliance verification on the scheme, generating evaluation data, and adjusting the scheme according to the evaluation data. And finally, performing space efficiency deduction calculation on the adjusted scheme, optimizing the scheme according to deduction result data, and outputting a final planning scheme. According to the method, the accuracy of data mining is improved through target-driven element extraction, and the perspectiveness and scientificity of a planning scheme are enhanced through dynamic efficiency simulation.

The invention discloses a scene perception information fusion decision-making method for industrial field multi-source heterogeneous data, which comprises an edge layer, a cloud layer and an interaction layer which form a three-level collaborative architecture, and belongs to the technical field of industrial field data processing. The method comprises the specific steps that S1, an edge layer collects industrial field multi-source heterogeneous original data and carries out self-adaptive preprocessing to obtain cleaned multi-source feature data; s2, the edge layer processes the multi-source feature data through a lightweight scene perception model, and outputs a preliminary scene judgment result; s3, the cloud layer constructs and updates a causal relationship graph based on industrial business knowledge and incremental data of the edge layer, processes data uploaded by the edge layer through a causal reasoning fusion model, outputs a decision result and issues the decision result through the interaction layer; and S4, the cloud layer optimizes parameters of the causal reasoning fusion model and the lightweight scene perception model based on a decision execution result fed back by the edge layer, and issues the optimized parameters to the edge layer through the interaction layer to realize dynamic iteration.

The invention discloses a returned intelligent electric meter sorting assembly line monitoring system. The system comprises a feature fusion sorting module, a digital mapping control module, a chain type tracing management module, a state diagnosis module and an environment management and control module. The feature fusion sorting module collects ammeter data through an industrial camera array, a microphone array, a structured light scanner and an infrared thermal imager, and performs feature fusion classification; the digital mapping control module calculates sorting process control parameters by using a distributed control model and an optimization algorithm; the chain type traceability management module adopts a block chain technology to record data and realize information retrieval; the state diagnosis module performs equipment fault prediction in combination with a data prediction technology; the environment management and control module monitors key parameters of the sorting environment. According to the system, the automation and intelligence level of an electricity meter sorting assembly line is improved, and the problems that an existing system is low in sorting efficiency, lagged in equipment monitoring and the like are solved.

The invention discloses an intelligent recognition method for a super-heavy beam under a multi-working-condition dynamic condition. The intelligent recognition method mainly comprises the following steps: defining a filter selection element; collecting and processing geometric data of the beam; detecting an intersection wall based on volume collision; through a multi-iteration combination algorithm, screening, classifying and identifying a beam upper wall, a beam lower wall and a beam side structure modeling line; further performing dynamic classification according to whether a beam side has a modeling line, whether a beam wall has a modeling line and a plate, and relative position relations thereof; defining corresponding calculation rules for different types of beams based on specifications and standards, and marking overweight beams; and automatic recognition of the overweight beam is completed. According to the method, automatic recognition work of the overweight beam under the complex actual engineering condition is achieved.

The invention discloses ultrapure water separation equipment and method based on use of a photo-thermal power plant, and particularly relates to the technical field of ultrapure water processing, the ultrapure water separation equipment based on use of the photo-thermal power plant comprises ultrapure water equipment, and further comprises a raw water tank arranged on one side of the ultrapure water equipment. Multi-point synchronous release is completed through the multiple stirring units, inert gas is promoted to enter raw water to complete replacement, control over the conductivity of the raw water is facilitated, multi-mode treatment of the raw water can be completed, the treatment precision and the treatment efficiency are improved, and the treatment cost is reduced. The ventilation mechanism and the discharging mechanism can be synchronously controlled through the control mechanism during operation, dynamic connection of the ventilation mechanism and the discharging mechanism is achieved, the ventilation mechanism can complete one-time switching in the single treatment process, the discharging mechanism is controlled to be dynamically separated from the sealing ring, the load is effectively reduced, and the efficiency is improved. The equipment is more stable in operation, and the service life is prolonged.

The invention provides a product-based feedback information processing method and device, equipment and a storage medium, and relates to the field of artificial intelligence, in particular to the field of big data and natural language processing. According to the specific implementation scheme, feedback information of a user is obtained; wherein the feedback information represents the use condition of the product by the user; performing validity analysis on the feedback information according to a preset vector database to obtain an analysis result; wherein the preset vector database comprises multiple pieces of vector data, the vector data are preset invalid information, the validity analysis represents that the validity of the semantics of the feedback information is analyzed, and the analysis result represents that the feedback information is valid or invalid; if the analysis result shows that the feedback information is valid, determining an information category to which the feedback information belongs; wherein the information category represents the effect of the feedback information on product improvement. The feedback information of the user is automatically processed, and the classification efficiency and precision of the feedback information are improved.

The present invention relates to the technical field of train operation control, and provides a train information prompt method, device, electronic device and storage medium, wherein the method comprises: obtaining the position information and train operation plan of a train to be processed, wherein the position information is determined based on a positioning device installed on the train to be processed; based on the position information and the train operation plan, selecting a candidate train corresponding to a target platform from the train to be processed; based on the position information of the candidate train and the position information of the target platform, providing information prompts on the position relationship between the candidate train and the target platform. The train information prompt method, device, electronic device and storage medium provided by the present invention can significantly improve the accuracy of PIS and PA information processing, reduce errors, and thus effectively improve the passenger experience.

The invention discloses a laying hen lossless data processing method based on improved XGBoost, and relates to the technical field of data processing, discrete feature numeralization is realized through adaptive tag coding, abnormal value detection and non-linear mapping correction are combined, a robustness purification feature set is constructed, and a layer lossless data processing method based on the improved XGBoost is obtained. The problems that a traditional method is sensitive to sensor noisy points and lacks an effective data cleaning mechanism are solved, and prediction precision reduction caused by extreme sample interference is avoided. And performing nonlinear expansion on the purification features by using a polynomial kernel, and accurately capturing a coupling relationship between the features. The global optimal hyper-parameter optimization of the XGBoost integrated regression model is driven by high-dimensional features, so that the parameter optimization efficiency is improved, local optimum is avoided, the method adapts to a multi-source heterogeneous and high-noise nonlinear data scene of laying hen breeding, the processing precision and generalization ability of the model are remarkably improved, reliable data support is provided for accurate management and quality monitoring of laying hen breeding, and the method is suitable for large-scale popularization and application. And efficient landing of the intelligent breeding technology is promoted.

The invention discloses an image acquisition method which comprises the following steps: S1, a coke pushing stage: uniformly increasing the air inlet speed of a cooling tower, reducing the air inlet temperature of the cooling tower, enabling the angle interval of a forward inclination angle a to be more than 0 degree and less than 90 degrees, and enabling a pushing unit to drive the cooling tower on a coke pushing rod to move for one forward step pitch at a time; continuously increasing the air inlet speed of the cooling tower to increase the rotation driving acting force of the acquisition lens, wherein the backward acquisition angle b is 90 degrees; according to the method, the acquisition angle of the lens is improved, under the condition of forward step pitch start-stop circulating motion, small step pitch motion is adopted in the coke pushing stage, meanwhile, the three-dimensional edge expressive force of an image acquired at a small inclination angle is higher, and under the condition that the lens and an algorithm are not changed, subtle furnace wall three-dimensional shape change can still be found; therefore, the image processing precision is indirectly improved. And in the returning state, the large-step-pitch returning step pitch is adopted for resetting.

The application provides a processing data table generation method and device, electronic equipment and storage medium, and belongs to the technical field of data processing. The method comprises the following steps: obtaining a target folder specified by a user and containing a plurality of structured data files, and creating a structured data table for each file; sequentially performing text column splitting, character cleaning, numerical value standardization and sorting on the structured data in each table to obtain a plurality of standardized data tables; extracting product processing data from each standardized data table according to a first target condition preset by the user and summarizing the product processing data to a target data table; sequentially performing screening matching, data deduplication and sorting on the data in the target data table to obtain a first standardized processing data table; and converting the table into a corresponding template file based on a target equipment type selected by the user and saving the template file to a specified path. The processing data table generation method and device, electronic equipment and storage medium provided by the application can improve batch data processing efficiency and ensure data whole-process accuracy.

The application provides a kind of wind power equipment multi-factor coupling data processing method, device and equipment, it is related to wind power generation technical field.The method comprises: collecting the original data set of wind power equipment;Based on the preset differentiating stratified pretreatment strategy, the original data set is carried out noise reduction processing, and the standard data set is obtained.According to the original data set, the correlation matrix of multi-factor coupling is constructed;Correlation matrix represents the coupling correlation degree between each data in original data set.According to the preset fusion model, the attention weight of each data in the original data set is updated, and each data is weighted and fused based on the updated attention weight, and the comprehensive feature set is output.According to the comprehensive feature set, determine the preliminary analysis data;According to correlation matrix, the preliminary analysis data is corrected, and the target analysis data is generated.The method is used to improve the data processing accuracy under the effect of multi-factor coupling interference.

The application discloses a data clustering method and device based on density peaks, a medium, and direct subordination is used to describe the relative density of each data point, wherein the data point is a picture or a word feature in a data set, a mixed density value combining the relative density and the absolute density is designed to identify a clustering center, and the effective subordination is defined through similarity reachability, main distribution of non-center points is performed, and the label of a remaining point is determined in combination with the k-layer upper distribution of different clusters, so that the data points of the same category in the data set can be effectively gathered in one category, and the classification of the data set is efficiently and accurately completed.

The present application relates to the technical field of simulation sandbox, and particularly relates to a simulation sandbox data base construction method and device, wherein the method constructs compatible domain object models in the localization BIM forward design stage; according to unstructured drawings, localization BIM models and other data types, SHP data, geometric feature data and structured topology data are extracted or generated; simulation element data is obtained through field-level direct mapping, SHP data is converted twice and coordinates are unified, and then geometric feature completion, line segment merging and topology reconstruction are performed, object construction and labeling are completed based on the simulation element data, and simulation sandbox data base is obtained. The present application solves the problems of low construction efficiency, poor compatibility and insufficient geometric accuracy, realizes standardized data processing and efficient construction, can be directly loaded into a simulation sandbox, is suitable for airport and other scenes, and improves simulation accuracy and reliability.

The present invention discloses an optical property modeling method and device, which includes: dividing a microstructure in a periodic medium into N layers of thin slices in the z direction; performing the following processing for thin slices of the same layer of microstructure: obtaining each closed area of ​​the thin slice projected on the periodic space xy plane, and representing the boundaries of each closed area as each polygon; obtaining the x coordinate and y coordinate of each polygon vertex, and the line segment features of each polygon, dividing the periodic space according to the line segment features, and obtaining the integral sub-intervals divided in the x direction and the integral sub-intervals divided in the y direction; calculating the Fourier coefficients and Toeplitz matrix of the dielectric coefficient of the periodic medium based on the integral sub-intervals in the x and y directions, and performing strict coupled wave analysis to achieve optical property modeling of the periodic medium. The method optimizes the division method of the integral interval, and efficiently and accurately obtains the Fourier coefficients and Toeplitz matrix of the dielectric coefficient of the periodic medium.

The invention provides a sintering dynamic knowledge graph construction method, and relates to the technical field of knowledge graphs, and the method comprises the steps: collecting full-chain data in a real-time sintering process, and carrying out the parameter classification, and obtaining a sintering target parameter and a sintering process parameter; carrying out trend relation analysis based on a preset data analysis method so as to establish a sintering trend influence knowledge graph; performing parameter classification on the sintering process parameters based on the target ore blending cycle, and matching a corresponding machine learning algorithm to construct a sintering target parameter prediction model; obtaining a sintering trend influence knowledge graph corresponding to a first historical data point having the highest goodness of fit with the real-time ore blending structure, and obtaining a second historical data point having the highest goodness of fit with the current production state based on the sintering target parameter prediction model to obtain a sintering quantization influence knowledge graph; and determining a sintering trend corresponding to the real-time ore blending structure and a quantitative influence knowledge graph result. Fusion of mechanism data can be realized, and system processing efficiency and precision are improved.