30results about How to "Realize the predictive function" patented technology

The invention discloses a diabetic patient assessment and management system. The system comprises a patient information collection module, a regression model establishment module, a data processing module, a regression coefficient calculation module, a risk grade assessment module and a follow-up visit setting module. According to the system, risk factors leading to the complication of a patient are found out through Logistic regression analysis, the risk factors are assigned with corresponding scores according to the risk degrees of the risk factors, the symptom of the patient is objectively scored through accumulation of the risk factors, and the degree of the complication risk is judged according to the score accumulation condition, wherein the score changes dynamically along with the change of the patient's condition. Meanwhile, as all data is stored in an internet server, as long as a computer capable of having access to networks is available, a doctor can complete the function anywhere by accessing a website with accessing software.

The invention belongs to the field of carbon emission inversion, and particularly relates to a carbon emission inversion system and method based on deep learning. The system comprises a server and a client, wherein the server firstly collects enterprise real-time carbon emission data and real-time power utilization data; then a real-time emission condition is calculated through a carbon emission calculation module; meanwhile, a carbon emission prediction module is used for predicting a future emission condition; a data abnormity monitoring module is used for detecting and analyzing abnormal conditions in a prediction result; and finally, the client is used for displaying the data collected by the server, a carbon emission real-time calculation result and a abnormal condition analysis result in real time. The system and the method have the advantages that the problem that enterprise carbon emission data cannot be displayed in real time through a three-dimensional BIM model in the prior art can be solved, abnormal data of carbon emission can be detected, and generation reasons can be analyzed.

The invention relates to a driver fatigue detection system and belongs to the technical field of detection equipment. The driver fatigue detection system comprises a CCD (charge coupled device) camera connected with a DSP (digital signal processor) through a video input decoder, and detection images are subjected to mode identification and processing. An infrared light source of the CCD camera is composed of two groups of infrared diodes, and the infrared diodes are uniformly distributed on a same ring of the same plane and same axis and are optionally switched. The infrared diodes emit infrared light of 850nm and 950nm in wavelength. The driver fatigue detection system is characterized in that judgment of lip syncing is added on the basis of human eye area judgment, so that probability of judgment missing or mistakes is decreased. By a method for differentiating odd-even frames of the images, positions and characteristics of eyes and lips can be accurately detected, and fatigue degree of a driver can be accurately detected. By the aid of a Kalman filter and a Mean-shift algorithm, the positions and the characteristics of the eyes and the lips are continuously detected; by a short-time tracking strategy, the eyes and the lips are tracked, and prediction is realized.

The invention relates to the field of informatization technologies and risk management and control forecasting and early warning, in particular to an enterprise safety risk early warning method based on an analytic hierarchy process and a grey theory. By means of quantification and indexation of each safety production correlation index, weight coefficient of each index is calculated or obtained by the analytic hierarchy process or an enterprise self-defined method, and enterprise safety production risk indexes of next stage are forecasted according to input of historical data of each enterprise index and with the grey theory in fuzzy mathematics as a modeling basis. The enterprise safety risk early warning method based on the analytic hierarchy process and the grey theory has the advantages that the historical data are used as the basis, after the modeling basis is determined, a risk early warning index model is subjected to mathematical verification and historical data verification, a model is properly modified, threshold value determining is performed on a production accident risk early warning model, immediate output of early warning signals is achieved so that safety conditions and future development trends in the enterprise safety production process can be visually displayed, and a forecasting function of enterprise safety production risks can be achieved.

The invention discloses a network attack target prediction method based on neighbor similarity. In the existing threat prediction methods, prediction and analysis are performed based on attack behaviors, and an attack target is not further predicted. The technical scheme adopted in the invention is as follows: firstly, preprocessing a safety event, performing normalization processing, and removing redundancy and misinformation; then matching the preprocessed safety event with a pre-defined rule library, performing correlation analysis, and reconstructing an attack scene; and finally calculating the similarity of a host address, an open port and an operating system of the attack target with these attributes of a neighbor host, and predicting the next step network attack target. The network attack target prediction method disclosed by the invention provides reference for an administrator to prepare handling strategies, achieves a prediction function of network attacks and improves the overall safety of the network.

The invention relates to the field of artificial intelligence, discloses a traffic jam prediction method, device and equipment and a storage medium. The method is used for solving the technical problem that a traffic jam condition cannot be predicted in an existing urban intelligent traffic system. The method comprises the following steps: according to a traffic jam prediction request, obtaining traffic flow data at each traffic gate to obtain a traffic flow data set; analyzing the parameters of the traffic checkpoints in the traffic flow data set by using a big data analysis technology to obtain node parameters; calculating the relationship among the traffic checkpoints to obtain edge parameters; and generating a traffic map in a preset time period based on the node parameters and the edge parameters through a pre-constructed road network knowledge graph, and inputting the traffic map into a traffic prediction model pre-established based on a graph convolutional neural network for road condition prediction processing to obtain a predicted value of the congestion degree of each traffic gate corresponding to the traffic congestion prediction request. In addition, the invention also relates to a block chain technology, and the related information of the traffic flow data can be stored in a block chain.

The invention provides a method and device for predicting fault of a turnout switch machine. The method comprises the following steps: acquiring action current data of a turnout switch machine and taking the current data as a prediction set; extracting current characteristic values of each current data in the prediction set; respectively carrying out decision and classification on the current characteristic value of each current data in the prediction set by adopting the decision tree of each working stage of the turnout switch machine to obtain a classification result of whether each currentdata in the prediction set is in fault or not in each working stage of the turnout switch machine, determining an abnormal current characteristic value when a certain current data in the prediction set is in a fault in any working stage of the turnout switch machine to obtain a fault prediction result. The decision tree of each working stage of the turnout switch machine is pre-established based on the obtained turnout switch machine action current data sample of each working stage, a characteristic value set of the current of the turnout switch machine in each working stage and a pre-established fault characteristic set of the turnout switch machine. The method and device for predicting fault of the turnout switch machine can realize the fault prediction function of the turnout switch machine.

The invention relates to a heat exchanger filth blockage detection method and device. The method comprises the steps of collecting the actual operation data of a heat exchanger during operation, and obtaining the theoretical operation state data through calculation according to the actual operation data and a predetermined operation state model, wherein the operation state model is determined according to operation data and operation state data when the heat exchanger is not blocked by dirt; acquiring actual operation state data corresponding to the actual operation data when the heat exchanger operates; and according to the actual operation state data and the theoretical operation state data, determining whether filth blockage occurs to the heat exchanger. According to the method, the filth blockage degree of the unit heat exchanger can be judged based on historical data, the judgment effect and the detection speed are improved, the problems of too large algorithm error and insufficient stability caused by few monitored operating parameters are solved, the accuracy of filth blockage judgment of the heat exchanger is improved, and the development trend can be summarized through historical operation data and real-time operation data, so that the filth blockage condition of the heat exchanger in the future is accurately predicted.

The invention belongs to the technical field of power grid engineering and in particular relates to a method for constructing a distribution network engineering cost data analysis model. The method includes the following steps of: 1) establishing a distribution network engineering cost data storage system based on a web service platform, wherein the storage system a distribution network engineering cost analysis system data module; 2) establishing a cost data element system within the cost data storage system; and 3) establishing a distribution network cost index mathematical model by using aBP neural network in a big data analysis method. The method, by establishing the distribution network engineering cost data storage system, helps technicians quickly determine cost by the big data analysis method, and improves service analysis depth and efficiency of.

The present invention provides a news sub-event prediction method and apparatus based on the depth learning technology. The method comprises: step 1: carrying out representation learning on the sub-event by using the LSTM depth learning technology based on the pre-processed large-scale sub-event sequence data; Step 2, carrying out representation learning on the sub-event sequence based on the sub-event representation learning and sub-event subjects; and step 3, predicting the next sub-event by using the sub-event sequence representation obtained in the step 2. According to the method provided by the present invention, the sub-event prediction function can be effectively realized, the sub-event not existing in the training corpus can be predicted, and the better prediction effect can be obtained.

The invention provides a building inclination angle prediction method and system. The method comprises the following steps: reading and inputting an inclination angle measurement value into a data matrix; according to a matrix decomposition method and a gradient descent method, fitting the inclination angle measurement value in the data matrix to obtain inclination angle predication values corresponding to all points in the data matrix, and outputting the inclination angle predication values when the preset conditions are met; saving data of the wind direction and the wind speed at the points, corresponding to the inclination angle predication values which exceed the inclination angle threshold, in the data matrix to form an inclination-prone wind direction and wind speed data set; comparing the predication values of the wind direction and the wind speed in an environment where a building is positioned with the data of the wind direction and the wind speed at such points in the easy-inclination wind direction and wind speed data set; when the predication value falls into a preset range, sending an early warning signal. Through the adoption of the method and the system, the inclination angle of the building in the preset wind direction and wind speed conditions can be predicted, so that the wind direction and wind speed corresponding to the situation that the building is prone to inclination can be obtained, the warning function is achieved, and the building is prevented from collapsing.

The invention discloses an automatic blue algae monitoring and early warning method and an automatic blue algae monitoring and early warning system. The method comprises the following steps: shootinga certain image of blue algae in a water area in real time to obtain a corresponding blue algae image; corroding and expanding the blue algae image, and filling holes to obtain an RGB histogram of theblue algae image; converting the blue algae image into a gray-scale image, and solving an LBP map to obtain an LBP histogram; obtaining color features and texture features of the blue algae image; weighting and fusing the color features and the texture features at a decision level to form fused features; identifying a blue algae area of the blue algae image according to the fused features; calculating a pixel proportion of the blue algae area occupying the whole blue algae image; and judging a standard range of the blue algae image according to the pixel ratio, and sending out corresponding early warning information. The invention has a good application effect on multi-feature fusion under small samples, high-dimensional space and uncertainty, does not depend on the number and quality ofthe samples excessively, and realizes monitoring and early warning of the blue algae.

The invention provides a micro-service state detection method, a model training method, equipment and a storage medium, and the method comprises the steps: obtaining a plurality of index parameters of a micro-service for each micro-service under a micro-service architecture, inputting the plurality of index parameters to a micro-service state detection model, obtaining a detection result used for indicating whether the micro-service is abnormal, wherein the micro-service state detection model is obtained by training a back-propagation neural network model. The micro-service state detection model of the embodiment is used for predicting the health state of the current micro-service, and the prediction function of the health state of the micro-service is realized by analyzing multiple index parameters of the current micro-service, so that abnormal micro-services are isolated in advance, the stability and continuity of the whole application or service are ensured, and the service quality of the system is effectively improved.

The invention discloses a multi-dimensional high-precision track intelligent prediction method based on line segment clustering. The method includes discretizing the continuous track data; carrying out abrupt change longitude data processing, data cleaning and normalization processing; compressing track data by using a Douglas-Peucker algorithm; using a DBSCAN clustering algorithm for clustering flight paths, selecting flight path clusters corresponding to emergency situations according to different emergency situations under multi-dimensional factors, performing flight path prediction througha flight path prediction neural network model, and finishing a multi-dimensional high-precision flight path prediction task. According to the invention, the original flight path data is compressed, so that the calculation pressure is greatly reduced under the condition of reserving flight path characteristics, the operation time is shortened, and the operation efficiency is improved; a convolution and LSTM neural network model is adopted, convolution is used for feature extraction, and the track prediction precision of the LSTM neural network model can be improved.

The invention discloses a big data analysis and processing platform of an air quality monitoring system, which applies the big data technology to the air quality monitoring system instead of using a traditional database to store the air quality data and tedious software programming to realize data processing and display. The big data analysis and processing platform comprises a data acquisition module, a data preprocessing module, a data processing module, a data storage module and a data prediction module, and can realize real-time synchronous processing and storage of data of a plurality ofmonitoring terminals. The processed data are stored in the distributed database HBase. By pre-partitioning the HBase table and designing the key values of the data, the air quality data can be saved by multiple computers and the hot spots can be solved effectively. By inputting temperature, humidity, wind direction and air quality data of neighboring cities, the data prediction module can effectively predict air quality data, and save the input and forecast data.

The invention relates to a large-break-angle welding seam tracking and obstacle prediction system based on binocular four-line vision sensing, and is applied in the field of welding automation and welding seam tracking. The system is technically characterized by comprising a crossed four-line type laser visual system, a single-line type laser displacement sensing system, an intelligent visual processing computer system, two industrial wide-angle lenses, two dimming and filtering systems, a support, a box body and a cover plate. According to the system, a welding-seam-like track of a full-pose large-break-angle workpiece is obtained through the crossed four-line type laser vision system; through synchronous scanning laser rays of the intelligent visual processing computer system on the crossed four-line type laser visual system, shooting by the industrial lenses and processing of light rays by the dimming and filtering systems, the assembly clearance of the large-break-angle workpiece is obtained; meanwhile, transverse distance data are obtained through the single-line type laser displacement sensing system; and welding seam tracking and obstacle prediction of the large-break-angle workpiece are realized through data processing fusion of the welding-seam-like track of the full-pose large-break-angle workpiece by the intelligent visual processing computer system, the assembly clearance of the large-break-angle workpiece and the transverse distance data.

The invention relates to an automatic computing system for piece inventories, comprising a virtual inventory judging unit, an import declaration unit, a virtual inventory maintaining unit and an actual inventory maintaining unit, wherein the virtual inventory judging unit is used for judging whether a piece number in a piece customs entry document is larger than a virtual piece turnover allowance or not, and the virtual piece turnover allowance is the difference of a piece turnover amount and a virtual piece inventory in an electronic account book; the import declaration unit is used for declaring the piece customs entry document to a customs supervision system when the virtual inventory judging unit confirms that the piece number in the piece customs entry document is smaller than the virtual piece turnover allowance; the virtual inventory maintaining unit is used for checking the virtual piece inventory in the electronic account book according to the piece number in the piece customs entry document; and the actual inventory maintaining unit is used for checking an actual piece inventory in the electronic account book when receiving the import customs clearance feedback of the customs supervision system. The invention also provides a corresponding method. By adding the calculation of virtual inventory amount in the customs declaration link, the invention realizes the precognition function of inventory.

The invention discloses a method for designing a linear active disturbance rejection controller based on neural network prediction, which comprises the following steps of: 1, acquiring and preprocessing sample data, and constructing a sample training set and a sample test set; 2, constructing an ELMAN neural network prediction model by using the sample training set, and verifying the ELMAN neural network prediction model by using the sample test set; and 3, inputting the working condition data to the ELMAN neural network prediction model to obtain an output value to transform the extended state observer. Through the design of the linear active disturbance rejection controller, the control effect of the controlled object under various interferences is improved.

The invention discloses a network attack target prediction method based on neighbor similarity. In the existing threat prediction methods, prediction and analysis are performed based on attack behaviors, and an attack target is not further predicted. The technical scheme adopted in the invention is as follows: firstly, preprocessing a safety event, performing normalization processing, and removing redundancy and misinformation; then matching the preprocessed safety event with a pre-defined rule library, performing correlation analysis, and reconstructing an attack scene; and finally calculating the similarity of a host address, an open port and an operating system of the attack target with these attributes of a neighbor host, and predicting the next step network attack target. The network attack target prediction method disclosed by the invention provides reference for an administrator to prepare handling strategies, achieves a prediction function of network attacks and improves the overall safety of the network.

The invention provides a time sequence signal prediction method based on a quantum filtering model, and the method comprises the steps: setting a filtering parameter value and an initial value of each function, obtaining all time sequence signal data, sequentially calculating a difference function at each moment, updating a weight function, and calculating a potential energy function. Calculating a wave function and filtered time sequence signal data according to a Schrodinger equation in a difference form, calculating the value of each quantile of the density matrix, carrying out regression on each quantile, predicting the quantile value of each moment, and sequentially predicting the time sequence signal data value of the next moment; and the time sequence signal data with the time greater than the last moment in the time sequence signal data is predicted. According to the method, on the basis of a basic quantum filtering model, filtering processing can be achieved without making any assumption processing on the distribution form of noise signals, and the prediction function of quantum filtering is achieved.

The invention discloses an experimental method for predicting quality of cokes at different height positions in a coke oven. The method includes steps: acquiring a height H from a coal loading port toa coking chamber of a to-be-tested coke oven, a conventional coal loading height h and an average width d of the coking chamber; determining a height h and a width d of an experimental coal container; determining a distance h' between an actual coal loading port and a coal line after coal leveling in an experimental coal container production coke oven; acquiring a total loading bulk density rho in the experimental coal container; determining a height h1 between a researched position and the bottom to obtain bulk density rho 1 at the h1; performing a coking experiment, accurately controlling bulk density of each section and the total loading bulk density rho to be identical, and establishing a corresponding relation between experimental coke and produced coke; after bulk density of each section is controlled to be identical to the bulk density rho 1 at the h1, acquiring a strength index of the experimental coke, and finally acquiring an index of coke produced at the height h1 between the researched position and the bottom according to the corresponding relation between experimental coke and produced coke.

The invention relates to a public opinion information-based product prediction method and system. The method comprises the following steps: extracting public opinion corpus related to set category products from public opinion data; extracting an entity from the public opinion corpus through an entity recognition model; counting the volume of a product entity in the entities in the public opinion data; adopting a text sentiment classification model to carry out sentiment discrimination on the product entity to obtain entity sentiment information; and calculating a hotspot score of the product entity according to the volume of the product entity and the entity emotion information. According to the technical scheme provided by the invention, the entity corresponding to the product and the emotional information and the volume of the entity are obtained by analyzing and identifying the public opinion data, and the hotspot score obtained by calculating the two aspects of information including the emotional information and the volume can reflect the possibility that the product becomes a hotspot product, so that the function of predicting the hotspot product is realized; and product enterprises can be helped to quickly track market trends, grasp business opportunities and closely follow public opinion wind directions.

The invention discloses an experimental method for predicting quality of cokes at different height positions in a coke oven. The method includes steps: acquiring a height H from a coal loading port toa coking chamber of a to-be-tested coke oven, a conventional coal loading height h and an average width d of the coking chamber; determining a height h and a width d of an experimental coal container; determining a distance h' between an actual coal loading port and a coal line after coal leveling in an experimental coal container production coke oven; acquiring a total loading bulk density rho in the experimental coal container; determining a height h1 between a researched position and the bottom to obtain bulk density rho 1 at the h1; performing a coking experiment, accurately controlling bulk density of each section and the total loading bulk density rho to be identical, and establishing a corresponding relation between experimental coke and produced coke; after bulk density of each section is controlled to be identical to the bulk density rho 1 at the h1, acquiring a strength index of the experimental coke, and finally acquiring an index of coke produced at the height h1 between the researched position and the bottom according to the corresponding relation between experimental coke and produced coke.

The invention provides a method and system for predicting the inclination angle of a building. The method includes the following steps: reading the measured value of the tilt angle and importing it into the data matrix; using the matrix decomposition method and the gradient descent method to fit the measured value of the tilt angle of the data matrix to obtain the corresponding points in the data matrix The predicted value of the tilt angle, and when the preset conditions are met, the predicted value of the tilt angle is output; all the predicted values ​​of the tilt angle exceeding the preset tilt angle threshold are saved in the wind direction and wind speed at the corresponding point in the data matrix, forming an easy-to-tilt wind direction Wind speed data set; compare the forecast value of the wind direction and wind speed of the environment where the building is located with the wind direction and wind speed at each point in the easily inclined wind direction and wind speed data set; when it falls within the preset range, an early warning signal is issued. It can predict the inclination angle of the building at a given wind direction and wind speed, so as to obtain the corresponding wind direction and wind speed when the building is prone to inclination, realize the alarm function, and prevent the building from collapsing.

The invention provides an electrical phase prediction device based on a second-order derivative algorithm, which belongs to the technical field of electrical comprehensive protection and aims to solve the problem of inaccurate switching of dual power supplies with small phase difference due to the action time of the actuator; the invention includes a power supply module. , a signal processing module, a hardware zero-crossing module, a phase monitoring module, a data fusion module, a phase prediction module, a communication module and a functional additional module; the design in the present invention can realize the phase prediction function and perform the protection action in advance for the comprehensive protection. Data base; avoids the risk of power failure when switching between dual power sources without considering the actuation time of the actuator.

The invention discloses an ontology-based medical dispute case public opinion early warning level prediction method, which comprises the steps of firstly constructing an ontology of a medical disputecase, and then predicting a public opinion early warning level, and the public opinion early warning level prediction comprises the following steps: firstly training a prediction model of the public opinion early warning level by using a word coding algorithm; training a prediction model of the public opinion early warning level by using a sentence coding algorithm; and finally, case element allocation weights in the ontology structure are combined with the word codes and the sentence codes to predict the public opinion early warning level. The method has the advantages that more time can be bought for a court to process public opinions, the efficiency of workers is improved, the negative influence of network public opinion crisis is eliminated, and judicial credibility is improved; constructing a medical dispute case ontology by using ontology knowledge, and semantizing case elements by using an ontology reasoning method; a machine learning algorithm is combined with an ontology structure to form a prediction model of a medical dispute case public opinion early warning level.

The invention provides an electrical phase prediction device based on a second derivative algorithm, belongs to the technical field of electrical comprehensive protection, and aims to solve the problem of inaccurate small-phase-difference dual-power-supply switching caused by action time of an actuating mechanism. The device comprises a power supply module, a signal processing module, a hardware zero-crossing module, a phase monitoring module, a data fusion module, a phase prediction module, a communication module and a function additional module. According to the design in the invention, a phase prediction function can be realized, and a data basis is provided for comprehensive protection to carry out protection actions in advance; and the risk of power failure during dual-power switchingdue to the fact that the action time of an executing mechanism is not considered can be avoided.

The invention provides a method and device for predicting fault of a turnout switch machine. The method comprises the following steps: acquiring action current data of a turnout switch machine and taking the current data as a prediction set; extracting current characteristic values of each current data in the prediction set; respectively carrying out decision and classification on the current characteristic value of each current data in the prediction set by adopting the decision tree of each working stage of the turnout switch machine to obtain a classification result of whether each currentdata in the prediction set is in fault or not in each working stage of the turnout switch machine, determining an abnormal current characteristic value when a certain current data in the prediction set is in a fault in any working stage of the turnout switch machine to obtain a fault prediction result. The decision tree of each working stage of the turnout switch machine is pre-established based on the obtained turnout switch machine action current data sample of each working stage, a characteristic value set of the current of the turnout switch machine in each working stage and a pre-established fault characteristic set of the turnout switch machine. The method and device for predicting fault of the turnout switch machine can realize the fault prediction function of the turnout switch machine.

The invention discloses a code statement generation method, device and equipment and a readable storage medium, and the method comprises the following steps: determining a conversion strategy for converting a random forest model according to the model type of the random forest model; if the conversion strategy is a single-layer strategy, converting decision trees in the random forest model into character strings, and generating a first prediction code statement according to each character string; if the conversion strategy is a nested strategy, converting decision trees in the random forest model into character string arrays; and according to each character string array, generating a probability prediction code corresponding to the decision category of each decision tree, and generating each probability prediction code into a second prediction code statement. According to the method, the random forest model is converted into the code statement, so that the random forest model and the code statement are compatible in service deployment.