76results about How to "Strong correlation" patented technology

The invention discloses a transmission network fault positioning method, system and device and a storage medium, and relates to the field of communication, and the method comprises the steps: firstly, collecting network data to construct a multi-granularity network topology model, then obtaining original alarm data, analyzing the topology association relationship between alarms, and obtaining a multi-granularity network topology model; and mining horizontal peer association and vertical upstream and downstream association among the alarm objects, and identifying multi-alarm data of each alarm object. Thirdly, constructing an associated alarm directed graph based on the multi-alarm data, and performing associated alarm compression on the associated alarm directed graph to obtain a compressed alarm set; and finally, constructing an alarm propagation graph based on the compressed alarm set and the multi-granularity network topology model, querying the alarm propagation graph through a reverse hierarchical sorting algorithm, and determining a root alarm. According to the method, through innovative methods such as topological modeling, association mining, graph compression and reverse sequencing, alarm data and a topological structure are closely combined, an alarm propagation rule is analyzed, and fault root causes are diagnosed.

The invention relates to the technical field of side slope monitoring and early warning, in particular to a mine site side slope remote online real-time monitoring and early warning system which comprises a multi-source data monitoring unit, a data processing and analyzing unit, a side slope instability judging unit, an instability grading early warning unit and an early warning verification pushing unit. The method comprises the following steps: collecting monitoring data of a plurality of data sources to obtain time sequence monitoring data of each data source, carrying out noise reduction on the time sequence monitoring data, then decomposing the time sequence monitoring data, calculating a time-frequency feature vector through an energy density function, and optimizing the time-frequency feature vector through feature screening; calculating a slope stability evaluation index by optimizing the time-frequency feature vector, constructing an instability risk index to judge a slope instability risk level, constructing an early warning index, and combining the early warning index with the slope instability risk level to establish a graded early warning mechanism to obtain a slope instability early warning level; and verifying the slope instability early warning grade, and performing early warning according to the slope instability early warning grade passing the verification. According to the invention, the false alarm rate of slope monitoring and early warning can be reduced.

The invention belongs to the technical field of intelligent questioning and answering, and particularly discloses a business questioning and answering method and system based on a large language model and a scene label, corresponding knowledge base question sets are associated through business question expansion matching, and then answers are called from a knowledge base to determine corresponding data blocks and scene labels. Then, through scene label matching and scene consistency evaluation, a database with high scene consistency is selected to be combined with user questions to be input into a large language model to achieve intelligent question and answer. The business question and answer requirements of the user can be effectively met, and the user does not need to directly query professional knowledge; according to the method, questions proposed by users can be analyzed more comprehensively, questioning intentions and question scene information are fully covered, and question and answer feedback results are more targeted and have higher relevance with business questions of the users; and the question and answer result is simpler and more effective.

The invention discloses a voice-driven facial animation simulation method based on facial muscle linkage, and belongs to the field of artificial intelligence, and the method comprises the steps: S1, constructing a PPMF encoder; s2, constructing a decoder based on FDCP to decode FP features provided by the PPMF so as to obtain a facial animation; s3, training a 3D face animation framework DCPTalk driven by voice; s4, optimizing the model; and S5, quantitatively evaluating the model. According to the invention, a DCPTalk framework is provided, and the Mouth2Face is provided based on the linkage characteristic of the facial muscle group. Motion of the mouth has strong correlation with voice signals, and can be synthesized with sound track dynamics easily. In order to further enhance the details of the facial movement, the Refine Decoder is adopted to simulate the skin deformation of the surface layer to refine the facial animation. Inherent body characteristics and body characteristics related to facial muscle group movement are embedded into Mouth2Face, a personalized facial muscle control system is constructed, and meanwhile, an external driving signal is modulated by using a speaking style. Qualitative and quantitative experiments and user research show that the DCPTalk is superior to the most advanced method in the prior art.

The embodiment of the invention provides an insurance product recommendation method and device, electronic equipment and a storage medium, belongs to the technical field of artificial intelligence, and is suitable for the field of financial science and technology. The method comprises the following steps: acquiring original user behavior data, and performing activeness screening on the original user behavior data to obtain active user behavior data; obtaining target behavior data of the target object; performing user similarity calculation based on the active user behavior data and the target behavior data to obtain target user similarity data; performing user screening on the active user behavior data based on the target user similarity data to obtain target similar users; determining a target insurance product from the active user behavior data based on the target similar user; and performing product recommendation to the target object based on the target insurance product. According to the embodiment of the invention, recommendation errors caused by rare user behavior data can be effectively reduced, and the accuracy of insurance product recommendation is improved.

The invention relates to an interaction method and device, computer equipment and a storage medium. The method comprises the following steps: determining a target intention related to a user according to multi-modal information related to a target vehicle; and starting a dialogue service flow, the dialogue service flow comprising: sending an instruction corresponding to the target intention to an application to which the atomization service function for realizing the target intention belongs, so that the application calls the atomization service function for realizing the target intention according to the instruction.

The invention provides a visual question and answer method based on knowledge retrieval enhancement, and relates to the field of computer vision and natural language processing, and the method comprises the following steps: respectively encoding a multi-modal query and a knowledge text by using a multi-modal encoder; calculating the similarity between the multi-modal query and the feature vector of the knowledge text by using a vector dot product, and retrieving to obtain a plurality of related external knowledge; the method comprises the following steps of: extracting the most useful visual representation from a frozen image encoder by using a lightweight visual Transform module; using a T5 encoder to independently encode the joint input of a plurality of knowledge paragraphs, images and questions; the first N most representative marks are connected in series and input into a T5 decoder for joint decoding, and finally a correct answer is generated. According to the method, the problems of visual information loss and insufficient knowledge utilization in the existing method are solved, and the performance of the system in the knowledge-intensive visual question and answer task is remarkably improved.

The invention relates to the field of sedimentary stratigraphy, in particular to a method for analyzing a deep-time El Nino-Southern oscillation cycle based on a sea-lake facies sediment annual stripe layer. The method comprises the following steps: sampling and grinding a sheet, carrying out high-definition scanning, estimating a deposition rate, carrying out element plane distribution, carrying out principal component analysis on the content of each element, carrying out laser in-situ carbon and oxygen isotope test and time domain spectrum analysis on an annual stripe layer one by one, and finally analyzing the climatic change of the interannual scale in the geological historical period. According to the method, element data, the thickness of the annual lamina and the carbon and oxygen isotope of the annual lamina are analyzed and explained in a combined mode, relevance exists among the indexes, a new method is provided for identification and analysis of the deep-time El Nino-Southern oscillation cycle, a time scale with the year as the unit is established on the basis of the marine facies or lacustrine facies annual lamina, and the identification accuracy of the deep-time El Nino-Southern oscillation cycle is improved. And the climate fluctuation is subjected to high-precision analysis under the scale, so that the related technical blank is filled.

The invention relates to an interaction method and device, computer equipment and a storage medium. The method comprises the following steps: determining a target intention related to a user according to multi-modal information related to a target vehicle; and starting a dialogue service flow, wherein the dialogue service flow comprises calling an atomization service function for realizing the target intention.

The invention provides a stacking model-based obesity genetic risk model construction method, electronic equipment and a storage medium, the method comprises the following steps: obtaining a data set, the data set comprising genetic feature data, and the genetic feature data comprising single nucleotide polymorphism; screening the single nucleotide polymorphisms by using a random forest to obtain a first genetic feature set related to obesity genetic risks; screening the single nucleotide polymorphism by using a distributed gradient enhanced library model to obtain a second genetic feature set related to the obesity genetic risk; obtaining a significant genetic feature set according to the first genetic feature set and the second genetic feature set; constructing an obesity genetic risk model; according to the method, the model evaluation result of the obesity genetic risk model is obtained, the parameters of the obesity genetic risk model are optimized according to the model evaluation result, stacking training is carried out by integrating multiple basic learners and combining meta-classifier logistic regression, the advantages of different algorithms are effectively fused, and the method is suitable for large-scale popularization and application. And the accuracy and the stability of obesity genetic risk prediction are improved.

The invention discloses a depression detection method based on image-text social media analysis, relates to the technical field of automatic detection of depression, and solves the technical problem in the prior art that whether a user is a depression patient or not is detected by analyzing character and picture data published by the user on a social media platform. Specifically, the method comprises four steps of data capture, post screening, image-text representation and fusion classification. Generally speaking, the method comprises the following steps: firstly, capturing a posted document or posted post of a user in a certain time period from a social media platform, including characters and picture data in the posted document; then, performing depression related symptom prediction on the posted document, and screening out the most related K posts; then, respectively carrying out text and picture representation on the screened posts; and finally, fusing the characterization vectors of the two modes for depression detection.

The invention discloses an infrared image column stripe heterogeneity correction method based on an improved median histogram, and belongs to the field of infrared image heterogeneity correction. The method comprises the following steps: calculating a column histogram of a single-frame infrared image and a column mean value curve subjected to wavelet denoising according to an input infrared image; performing inversion operation according to the obtained column histogram to obtain a column inverse histogram; generating an adaptive window of the bilateral filter according to the column mean value curve; generating a bilateral filter according to the spatial domain distance of the column and the gray domain difference of the column mean value curve; filtering the column inverse histogram according to the obtained bilateral filter to obtain a median histogram; and an input infrared image is mapped according to the median histogram to obtain an infrared image after stripe removal. According to the method, a better fringe removal effect is achieved, blind pixel points are avoided, and the step of repeatedly calculating a comparison result to solve an optimal parameter value in an original algorithm is omitted.

The invention discloses a power transformer turn-to-turn short circuit fault positioning method and system based on magnetic flux leakage distribution. The method comprises the following steps: acquiring the radial magnetic field intensity inside a power transformer; based on the obtained radial magnetic field intensity, the short circuit fault occurrence condition is judged by adopting a difference value method; when a short-circuit fault occurs, the position of the short-circuit fault point is positioned by using the position of the zero crossing point of the radial magnetic field intensity. According to the invention, the improved optical fiber sensor is used for collecting the magnetic flux leakage distribution in the power transformer, and then the amplitude and phase characteristics of the magnetic flux leakage are obtained through Fourier transform and an interpolation method. And rapid monitoring and interval positioning of the turn-to-turn short circuit fault of the power transformer are realized by monitoring the internal magnetic leakage amplitude and phase characteristics of the power transformer.

The invention relates to the technical field of oil field oil production engineering, in particular to an underground data acquisition unit with flow, pressure and temperature monitoring functions and a method. The method is suitable for underground working condition monitoring of submersible motor type lifting equipment such as submersible screw pumps, submersible plunger pumps and submersible centrifugal pumps. The unit comprises a pressure-resistant shell, wherein a central pipe, an electromagnetic flowmeter, two pressure and temperature sensors and an underground parameter collector are integrated in the pressure-resistant shell; the two sensors are used for measuring the temperature and pressure of liquid flow of the central pipe and the temperature and pressure of an oil casing annulus respectively; the electromagnetic flowmeter measures the flow, and real-time temperature and pressure data are used for compensation to improve the precision. The unit is further provided with a collecting pipe, an upper end, a lower end, a connecting sliding ring, an annulus / oil pipe inner testing channel, a sealing rubber ring and the like. Through the integrated design, synchronous high-precision in-situ monitoring of underground key parameters is achieved, the problem that underground data of multiple instruments are not synchronous is solved, and a reliable data terminal is provided for an intelligent oil field.

The invention belongs to the technical field of waste heat recovery control, and provides a natural gas cold and heat cogeneration control method, which comprises the following steps: setting a monitoring period, monitoring two ends of each natural gas conveying pipeline to obtain natural gas temperature data, analyzing and processing the natural gas temperature data to obtain an ideal heat recovery value, and controlling the heat recovery value. Analyzing and processing the natural gas temperature data again to obtain an actual heat recovery value, acquiring a gas flow in each natural gas conveying pipeline in a monitoring time period, obtaining a gas flow mean value, processing the actual heat recovery value and a theoretical heat recovery value, outputting to obtain a heat recovery efficiency value, and comparing the heat recovery efficiency value with a threshold value; if the heat recovery efficiency value is smaller than the heat recovery efficiency threshold value, it is indicated that the heat recovery efficiency of the heat recovery system in actual operation is low, and a low recovery efficiency signal is generated, so that comprehensive evaluation on the heat recovery efficiency of the heat recovery system is completed by setting a monitoring period.

The invention provides a product classification model training and predicting method based on a digital human, and belongs to the technical field of product classification and digital human. The training method comprises a training corpus collection step, a training corpus classification step, a digital human generation step, a model training step and a digital human model association step. The prediction method comprises the following steps: S710, receiving information of a to-be-classified product input by a user; s720, analyzing the information of the to-be-classified products to obtain a plurality of semantic segmented words; s730, taking the semantic segmented words as the input of the target product classification model; and S740, outputting a product classification result through a digital human model associated with the target product classification model. According to the technical scheme provided by the invention, the output result of the target product classification model can be output through the digital human, so that the output information of the product classification result is richer and more visual, the generated digital human model and the product classification model also have relatively strong relevance, and the user experience is good.

The present application relates to the field of image processing technology, and in particular to a method and system for monitoring the ecological environment of a mine based on image processing, the method comprising: collecting monitoring data of each monitoring point in the monitoring mine at a historical moment; calculating the environmental complexity of each monitoring point and the local Moran index of multiple preset scales; calculating the selection score of each preset scale; selecting N preset scales according to the selection score, and counting the proportion of the number of monitoring points whose local Moran index is greater than the index threshold in the N preset scales to draw a proportion curve; determining the inflection point of the proportion curve, and using the preset scale corresponding to the inflection point as the target scale to construct a convolutional neural network, which can output the monitoring results of each monitoring point. Through the technical solution of the present application, the ecological environment monitoring results at various locations in the mine can be accurately obtained.

The application relates to a business processing method and device, computer equipment, a storage medium and a computer program product, and relates to the technical field of cloud computing. The method comprises the following steps: determining a target group according to attribute data of a target object; obtaining group correlation degrees of each control group and the target group from a group correlation graph; determining the correlation degrees of each second historical business and the target object according to the group correlation degrees of the target group and the control group and the business correlation degrees of each second historical business and each first historical business; and determining a target business according to the correlation degrees of each second historical business and the target object in each control group. By using the method, the control groups having the correlation with the target object and the historical business data of the control groups having the correlation with the historical business data of the target object are fully referenced, so that the target object can be recommended with the business having a stronger correlation with the historical business, and the accuracy and reliability of business recommendation are improved.

The application discloses an advertisement conversion rate calibration method and device, electronic equipment and a storage medium. The method comprises the following steps: obtaining a parameter cumulative value related to an estimated conversion rate of target advertisement data within a preset time length, each set of target advertisement data having the same dimension information under at least one dimension object and corresponding to the same identifier combination, and the target advertisement data being selected from M advertisement data; selecting a target parameter cumulative value from the parameter cumulative values of the multiple sets of target advertisement data corresponding to the preset time length according to the to-be-put advertisement data and the identifier combination corresponding to the to-be-put advertisement data; obtaining a calibration coefficient of the to-be-put advertisement data according to the target parameter cumulative value; and calibrating the estimated conversion rate of the to-be-put advertisement data by using the calibration coefficient to obtain a calibrated estimated conversion rate. By using the calibration method, the calibrated estimated conversion rate is more accurate.

The invention relates to the technical field of image processing, in particular to a mine ecological environment monitoring method and system based on image processing, and the method comprises the steps: collecting monitoring data of each monitoring point in a monitored mine at a historical moment; calculating environment complexity of each monitoring point and local Moran indexes of a plurality of preset scales; calculating a selection score of each preset scale; n preset scales are selected according to the selection scores, and the proportion of the number of the monitoring points with the local Moran indexes larger than an index threshold value in the N preset scales is counted to draw a proportion curve; an inflection point of the proportion curve is determined, a preset scale corresponding to the inflection point serves as a target scale to construct a convolutional neural network, and the convolutional neural network can output monitoring results of the monitoring points. According to the technical scheme, the ecological environment monitoring result of each position of the mine can be accurately obtained.

The invention relates to the technical field of tunnel engineering, in particular to a cold region operation tunnel lining frost heaving influence test device and method.The device comprises a walk-in test box used for providing a uniform and controllable low-temperature environment, and a model box is arranged in the walk-in test box; a surrounding rock similar material, a tunnel lining model and a ponding simulation mechanism for simulating a ponding structure are arranged in the model box, the tunnel lining model and the ponding simulation mechanism are surrounded by the surrounding rock similar material, stress loading mechanisms are arranged on the top and the side wall of the model box, and a detection mechanism for detecting stress deformation displacement of the tunnel lining model is further included. According to the test device and the test method adopting the test device, the whole model is uniformly and controllably cooled through the step-in test box, a real temperature field where the tunnel in the cold region is located is accurately simulated, and the problems of uneven temperature and frost heaving simulation distortion caused by local refrigeration in the prior art are solved.

The present invention relates to the technical field of fiber optic gyroscopes, and in particular to a method for preparing a two-input and two-output fiber optic ring, comprising the following steps: measuring two optical fibers of required lengths and reserving fiber pigtails of sufficient length; marking the end points and midpoints; using a quadrupole symmetrical winding method to alternately wind the two optical fibers on a fiber optic ring skeleton; changing the winding sequence of the two optical fibers; leading the four fiber pigtails out of the fiber optic ring skeleton and fixing them; and fixing the wound fiber optic ring. According to the method for preparing a two-input and two-output fiber optic ring of the present invention, firstly, the two optical fibers are alternately wound on the fiber optic ring skeleton, and the two optical fibers maintain physical consistency in space, thereby ensuring a strong correlation between the two optical fibers. Secondly, the two optical fibers are alternately wound on the fiber optic ring skeleton, achieving the equivalent effect of the correlation between the two channels of dual polarization while eliminating the polarization residual coherent phase error of the dual polarization in a single optical fiber.

The invention relates to the technical field of integrated navigation, in particular to a positioning method and system based on optical angle measurement and astronomical-inertial integrated navigation. According to the method, firstly, a star image and a navigational satellite image are obtained at the same time by shooting a star map, after recognition and processing, the attitude of a camera relative to a geocentric inertial coordinate system is calculated by utilizing the star, and then the coordinate of the navigational satellite in the coordinate system is determined. And obtaining angle information of the navigation satellite relative to the observer under the earth coordinate system through coordinate transformation. Optical angle measurement information is fused with attitude, speed and position information output by an inertial navigation system, a Kalman filtering state equation and an observation equation are constructed, and finally real-time high-precision calculation and updating of the position of the observer are achieved. According to the method, the accuracy of optical angle measurement, the limited distance reference of celestial navigation and the autonomous continuity of inertial navigation are fused, and the navigation positioning accuracy and reliability under the complex environment and the satellite denial condition are effectively improved.

The invention provides a multi-channel pneumatic heating test system and method, and relates to the field of aircraft testing. In the heating test system, a plurality of heaters and a plurality of temperature sensors are respectively connected with a controller, the heaters heat a tested area of the aircraft, the temperature sensors acquire the temperature of the tested area, and the controller can acquire the temperature from the plurality of temperature sensors in real time and determine the temperature change condition of first temperature data. According to the application, the temperatures of the plurality of tested areas of the aircraft can be synchronously tested, and the synchronous test of the plurality of tested areas of the aircraft can be realized, that is, the temperature changes of different to-be-tested areas can be obtained, and the heaters corresponding to the different to-be-tested areas can be respectively controlled, so that the execution of the temperature test of each area of the aircraft can be further realized.

This invention provides KASP primer pairs, kits, methods, and applications for detecting SNP combinations related to soybean protein content. The SNP combinations related to soybean protein content include SNP1 at position 4903434 on chromosome Chr11, SNP2 at position 47558915 on chromosome Chr10, and SNP3 at position 47559046. This invention also provides three sets of KASP primer pairs for detecting SNP1, SNP2, and SNP3, used to detect soybean genetic DNA and genotype the three SNP loci to determine whether the species has high protein content. The detection method provided by this invention is fast, low-cost, and suitable for large-scale sample testing. This detection method can be applied to soybean molecular marker-assisted breeding selection, variety improvement, and germplasm resource evaluation.

The embodiment of the invention provides an optical attenuation interval positioning method and device of an ODN network and electronic equipment, and relates to the technical field of communication, and the method comprises the steps: obtaining a first regression equation of the length of a line of a cluster of the ODN network and the whole-course actual optical attenuation value of the line of the cluster; determining a whole-course actual light decay value of one line, a whole-course virtual light decay value of the line calculated by the first regression equation of the cluster where the line is located, and a deviation value between the whole-course actual light decay value and the whole-course virtual light decay value; removing a part of lines from all lines of one cluster; obtaining a second regression equation of the length of the line of the cluster and the whole-course actual light attenuation value of the line of the cluster by adopting the line with the line removed from the cluster; acquiring a whole-course actual light attenuation value of the line of the cluster and a third regression equation of a deviation value between the whole-course actual light attenuation value of the line of the cluster and a whole-course virtual light attenuation value of the line of the cluster calculated by the second regression equation of the cluster; and on the basis of the third regression equation of the cluster, the light attenuation interval in the cluster is positioned, so that the light attenuation interval positioning is more accurate.

The invention relates to the technical field of material informatics, and discloses a material data management method and system based on artificial intelligence, and the method comprises the following steps: S1, carrying out the first-level classification according to the purposes corresponding to collected material data; s2, on the basis of the first-level classification, performing second-level classification on the material data according to the main performance of the material; s3, on the basis of the second-stage classification, performing third-stage classification on the material data according to the preparation process of the material; and S4, on the basis of the three-stage classification, performing four-stage classification on the material data according to the auxiliary performance of the material. Through four-level hierarchical association, a complete link of application-main performance-preparation process-auxiliary performance is established, data of ten types of polymer fine processing products is effectively planned and managed, and efficient support is provided for research and development links such as formula screening, process matching and performance prediction, so that the speed of research and development and iterative optimization of new materials is accelerated.

The invention discloses a load prediction method and device considering cumulative temperature effect and heat island effect coupling, and the method comprises the training and use of a load prediction model, and the training process comprises the steps: determining a heat island effect temperature and a heat island reference temperature in the temperatures of a plurality of urban sub-regions in a target time period, taking the difference value between the two as a heat island effect; determining a sampling moment when the rising amplitude of the urban load capacity is about to enter a rising trend, taking a corresponding target urban temperature as a sensitive temperature, and calculating a temperature accumulation effect in a target time period based on the sensitive temperature; and taking the accumulated temperature effect, the heat island effect, the target city temperature and the city load capacity as training data for training a load prediction model. By adopting the above technical scheme, the relevance between the temperature rise and the load data is strong especially in summer, the change trend of the load data can be reflected, and a more accurate prediction result of the load data can be obtained.

The invention discloses an outpatient department accurate shunting and appointment guiding method based on AI pre-inquiry analysis. The method comprises the following steps that S1, a patient initiates pre-inquiry: the patient initiates a pre-inquiry request through a hospital APP, a self-service machine or an applet terminal; and S2: pushing the standardized inquiry questionnaire: the hospital APP, the self-service machine or the applet end pushes the standardized inquiry questionnaire through the shunting reservation system. By integrating multi-dimensional information such as texts, images, audios, patient ages and medical histories and by means of BERT semantic understanding, a medical knowledge graph and an XGBoost algorithm, the department matching accuracy is larger than or equal to 95%, the wrong number hanging rate is reduced to 3% or below, the limitation of single symptom matching is broken through, the cross-specialist differential diagnosis ability is improved, AI model driving replaces a rigid rule base, and the accuracy of diagnosis is improved. And iterative optimization is carried out through 1-20 thousand newly added data in each month, and rapid updating of the adaptive medical knowledge is realized.

The invention provides a bus load prediction method in the travel peak period of holidays and festivals. The bus load prediction method comprises the steps of collecting historical passenger data entering and leaving a certain region and bus load historical data; the collected data are preprocessed, the passenger variation of each day is calculated, and the bus load variation of each day is calculated; dividing the passenger variable quantity and the corresponding bus load variable quantity into a training set and a verification set; constructing a first neural network model, training the first neural network model by adopting the training set, and adjusting a hidden layer weight parameter of the model; verifying the trained model by adopting a verification set; if the verification is passed, outputting a first standard model for predicting the bus load change in the tourism peak period of the region; if the verification is not passed, returning to the step S3, re-dividing the data set, and then training; and collecting and predicting passenger variation, inputting the passenger variation into the first standard model to calculate and predict the variation of the bus load of the power grid, and further calculating the bus load of the power grid according to the variation.