51results about How to "Fast and Accurate Prediction" patented technology

A short-term traffic flow forecasting method based on road network space relation constraint Lasso is characterized in that road network constraint is realized by the least absolute shrinkage and selection operator Lasso, road network short-term traffic flow is forecast in three steps of road network analysis, model training and real-time forecasting. The invention provides the short-term traffic flow forecasting method based on the road network space relation constraint Lasso, actually measured expressway traffic flow data are combined, an Lasso algorithm is utilized to forecast the short-term traffic flow, a road network correlation analysis and variable selection method is given, and the expressway short-term traffic flow can be forecast relatively, accurately and rapidly.

The invention discloses an optical flow estimation method based on multiple dimensioned corresponding structuring learning; under given successive video frames, the method can parse the motion conditions of the first frame relative to the second frame; the method specifically comprises the following steps: obtaining a successive frame image data set used for training optical flow estimation, and defining an algorithm object; carrying out structuring modeling for correspondences between two successive frame images in different dimensions; joint encoding the corresponding relations in different dimensions; building an optical flow estimation prediction model; using the prediction model to estimate the successive video frame optical flow values. The optical flow estimation method is applied to the optical flow motion analysis in true videos, and can provide better effect and robustness under various complex conditions.

The invention discloses a multi-modal trajectory prediction method for pedestrians in a complex scene. The method comprises the following steps: performing picture feature extraction by using a 16-layer convolutional neural network of a visual geometry group; performing feature processing on the trajectory data by using a full connection layer; inputting a trajectory data feature vector VS to enter a generative adversarial network to complete a coding and decoding network function; inputting picture feature data and track feature data to physics, wherein a social attention module considers terrain limitation and pedestrian interaction; obtaining a better track generation prediction result through the updated generator part; and obtaining a stable trajectory prediction model SPM. Accordingto the method, the prediction precision can be effectively improved, a plurality of reasonable prediction tracks can be generated, the related terrain limitation information can be extracted accordingto the feature information of the original picture, and the social interaction situation between different pedestrians in the same complex scene can be considered. The method can predict the future track of the pedestrian more quickly and accurately.

The invention provides an electric vehicle battery pack state-of-health prediction method and system based on big data. The method comprises the following steps: obtaining the data of a battery pack,and preprocessing the data to obtain preprocessed data; extracting characteristic parameters influencing the state of health of the battery pack in the preprocessed data, performing comparative analysis on the correlation between the characteristic parameters through a characteristic and target relation function, and deleting the characteristic parameters of which the correlation threshold is higher than a first threshold to obtain simplified model data; dividing data types of the data of the battery pack, expanding data with few orders of magnitudes in quick charging and slow charging, and marking battery quick charging time and battery slow charging time as supplementary characteristic data; establishing a battery state-of-health prediction model based on the preprocessed data, and / or the simplified model data, and / or the supplementary characteristic data, and adjusting parameters of the battery state-of-health prediction model; and calling the battery state-of-health prediction model through a flash framework, and calculating a prediction result of the state of health of the battery pack on line through a cloud network.

The invention discloses a real-time compensation method based on the motion prediction of a least squares support vector machine (LS-SVM) in a stable imaging system. The method comprises the followingsteps: firstly, establishing an LS-SVM motion prediction model and setting a parameter; then collecting motion detection data of the imaging system so as to train the motion prediction model; then, outputting a motion vector predicted value of a certain time in the future when a motion vector value of the current time is input; and finally, carrying out digital-to-analogue conversion on the motion vector predicted values by a stable compensation system to generate a compensation vector so as to control a compensation mechanism to carry out real-time compensation on a current motion of a camera. The invention can quickly and accurately predict the current motion state of the camera under the condition of an unknown camera motion model so as to achieve the purposes of detecting, delaying, predicting and compensating the motions, realize the real-time compensation of the motions of the camera, and improve the working performance of the stable imaging system.

The invention discloses an overall measurement method of pipeline vibration response, and relates to the field of engineering measurement. The method comprises the following steps: S1, establishing apipe model based on a pipeline system structure to be measured, and establishing a dynamic model based on the pipeline model; S2, arranging a plurality of measuring points for acquiring pipe wall vibration signals in a pipeline system to be measured; S3, acquiring pipe wall vibration acceleration signals of the measuring points through an acceleration sensor, and meanwhile converting an acquired time domain vibration acceleration signal into a frequency domain vibration acceleration signal through Fourier transform; S4, determining the position and size of an equivalent excitation source in the pipeline system according to the time domain vibration acceleration signal or the frequency domain vibration acceleration signal to obtain an equivalent excitation source vector; and S5, bringing the obtained excitation source vector into the dynamic model, and getting the vibration response situation of a position to be inspected on the pipeline in conjunction with a formula. Through adoption of the method, the vibration response situation of any point on the whole pipeline is accurately predicted through vibration measurement of a few measuring points.

The current invention concerns a method for determining the presence of a (pre)cancerous cell in a liquid cell sample, comprising the steps of: —suspending and preserving a cell sample obtained from a subject in a sample vial comprising a liquid medium, said liquid medium comprises means for labeling cells or epitope(s) on or in said cells; —obtaining data from said labelled liquid cell sample; and —determining the presence of said (pre)cancerous cells based on said obtained data; characterized in that said data comprises morphological data and biomarker data. In a further aspect, the invention relates to a liquid medium for fixing, preserving and labeling cells in a cell sample and a sample vial specifically designed to be used in conjunction with the current method.

The invention belongs to the technical field of electric welding quality control, and particularly relates to a resistance spot welding quality prediction method based on ensemble learning, which comprises the following steps: collecting welding process data of a welding spot sample according to process parameters measured by a sensor in a welding process; constructing a database; preprocessing the input data set through features; establishing an ensemble learning model for solder joint governance prediction, wherein each classifier outputs a quality prediction result of a to-be-detected sample; and according to output results of different classifiers and a voting mode, integrating results of the different classifiers for predicting the quality of the welding spot sample to be tested, andtaking most of judgment results as final prediction output. According to the method, the problems of high loss and low efficiency in the traditional welding spot quality detection process can be effectively solved, the welding spot quality can be quickly and accurately identified and predicted based on the welding process parameters, the welding spot quality analysis efficiency of electronic components is greatly improved, and the production cost is saved. The method is used for predicting the resistance spot welding quality.

The invention aims to provide a variable working condition emission performance prediction method for a combustion chamber of a gas turbine, and the method comprises the following steps: carrying outmesh generation and numerical simulation calculation of a combustion chamber model through FLUENT software for some typical working conditions; dividing the combustion chamber into five characteristic areas including a head mixing area, a main combustion area, a wall surface cooling area, a backflow area and a mixing area according to the combustion process, temperature distribution and flow field analysis; according to the divided feature regions, establishing a chemical reactor network (CRN) model for the combustion chamber of the gas turbine by utilizing Chemkin-Pro software; determining the effective volume of the feature region by adopting an infinitesimal method; and finding out calculation parameters of the reactor under other working conditions by adopting a Newton interpolation polynomial method, and predicting the emission performance under unknown working conditions. A method of combining fluent numerical simulation with a chemical reaction network is utilized, and the NOxemission performance of the combustion chamber of the gas turbine at different fuel-air ratios under the design working condition and the bias working condition is rapidly predicted.

The invention discloses a prediction method of aircraft scene taxiing time based on multiple regression analysis. Various factors affecting the aircraft scene taxiing time are comprehensively analyzed, according to different scene taxiing time data transformation ways and independent variable screening strategies, fundamental factors which affect the scene taxiing time are extracted, aircraft scene taxiing time prediction models based on multiple linear regression analysis and multiple curve regression analysis are established, and prediction errors of the established different models are comparatively analyzed, so as to achieve important significance for enhancing the prediction ability of the aircraft scene taxiing process in an airport flight area. The prediction method of the aircraft scene taxiing time based on the multiple regression analysis disclosed by the invention can accurately and fast predict the taxiing process of any approach and departure aircraft in pre-tactical and tactical levels, and can further provide a data reference basis for normal statistical standard formulation of strategic level flights, so that the problem of lower macro prediction ability in the airport aircraft taxiing process is effectively solved.

The invention relates to the technical field of computer vision, in particular to a video understanding method which is aimed at solving the technical problems of how to extract the dense frame feature and long-term spatiotemporal feature of a video effectively. For the aim, the video understanding method provided by the invention includes the steps that first, a plurality of video frame groups ofa target video is obtained by residual network; then time sequence relation network is used to generate the time sequence relation feature of the video according to the video frame groups; and finally, the video behavior category of the target video is predicted according to the time sequence relation feature, wherein each video frame group comprises two ordered video frames, each ordered video frame comprises a plurality of video frames arranged in chronological order in sequence. Based on the above steps, the dense frame feature and the long-term spatiotemporal feature of the target video can be effectively obtained so that the video behavior category of the target video can be predicted quickly and accurately.

The invention provides a prediction method, system and equipment for a scenic spot ticket quantity, and a storage medium. The method comprises the following steps that: dividing daily ticket quantityhistorical information into first ticket quantity historical information corresponding to statutory holiday vacations and second ticket quantity historical information corresponding to non-statutory holiday vacations; according to the second ticket quantity historical information, establishing a time sequence model; according to the first ticket quantity historical information, establishing a statutory holiday vacation ticket quantity prediction model; obtaining a ticket reservation quantity for a reference date by preset days before each reference date as a characteristic variable, and establishing a ticket quantity regression model according to the characteristic variable and the ticket quantity of the reference date; and selecting one of the time sequence model, a statutory holiday vacation ticket quantity prediction model and the ticket quantity regression model to carry out daily ticket quantity prediction. By use of the method, the historical tendency information of the scenic spot ticket and the reservation information, the daily ticket quantity of online travel agency tickets can be quickly and accurately predicted, and a guarantee is provided for reasonably arranging laborand guaranteeing user experience.

The present invention discloses a PHM platform. The PHM platform comprises: a modeling space unit. The modeling space unit comprises a data flow construction module, a flow management module, a flow verification module, a flow operation module and a console output module. A user uses the data flow construction module to acquire data so as to complete data flow construction by using a simple graphical operation manner of drag and drop and the connection; management operations and validity verification are performed on the data flow through the flow management module and the flow verification module so as to assist the user to complete the model operation and debugging; and a feedback result is provided to the console output module, and the execution feedback result of the flow is visually displayed so as to complete the entire modeling. The PHM platform further comprises a data management unit, an operation management unit, a system management unit, and a page display unit. According tothe principle of the present invention, monitoring of operating states, big data management and analysis mining, fault diagnosis and prediction of the to-be-detected equipment are implemented; the maintenance cost is low, the abnormality diagnosis is simple and real-time monitoring is achieved; and a theoretical basis is provided for military electronic product fault prediction capability.

Embodiments of the invention disclose a data processing method, a device, equipment and a storage medium. The method comprises the steps of obtaining a first camp in a game client and a service auxiliary parameter corresponding to the first camp; wherein the first camp comprises N to-be-competitive users; wherein N is a positive integer; determining timestamps based on camping to obtain game attribute characteristics of the N to-be-competitive users; when a target network model associated with the game client is obtained, determining a first group competition influence degree corresponding tothe first camp based on the game attribute characteristics of the N to-be-competitive users and the target network model; and when the first camp meets the business deadline condition, determining a second camp corresponding to a second group competition influence degree matched with the first group competition influence degree, and determining the second camp and the first camp as competition object groups corresponding to the game business. By adopting the method of the invention, the prediction efficiency and the prediction accuracy of the competitive influence degree can be improved, and then the camp matching accuracy can be improved.

The invention relates to a water quality biotoxicity prediction method based on an artificial neural network. The method comprises the steps: detecting a water body water sample through a bioelectrochemical water quality comprehensive toxicity sensor, and collecting the detected current change data ofa normal water body water sample and a toxic water body water sample; with the current inhibitingrate at a specific time node being input and the toxicity of the water being output, selecting a perceptron neural network containing three layers, wherein the number of hidden layers of the perceptron neural network is 1, and a system structure is self-defined; learning and training the neural network by means of batch processing training types and a conjugate gradient optimization algorithm, andfinally acquiring a neural network model capable of accurately predicting the water quality biotoxicity. Based on a bioelectrochemical water quality comprehensive toxicity sensor with high detectionsensitivity and an artificial neural network model with good prediction performance, the biotoxicity of the water quality of the water body can be rapidly and accurately predicted.

The invention discloses a method for designing components of a high-entropy alloy based on an electronic alloy theory. The method comprises the following steps: firstly, calculating an average Md value and an average Bo value of a target high-entropy alloy; then obtaining the phase structure of the target high-entropy alloy according to the relationship between the average Md value and the averageBo value, obtaining the weight percentage of each component according to the phase structure and the expression of the target high-entropy alloy, and finally weighing each component according to theweight percentage for alloying smelting to obtain the target high-entropy alloy; obtaining the phase structure of the target high-entropy alloy according to the relationship between the average Md value and the average Bo value, specifically, determining the phase structure of the target high-entropy alloy according to the relationship between the average Md value and the average Bo value. According to the method, the phase structure of the high-entropy alloy is predicted on the basis of the electronic alloy theory for the first time, prediction is rapid and accurate, great waste of cost and time due to the fact that the phase structure of the alloy can be known only by analyzing the alloy after actual smelting in the prior art is avoided, and the method has great application prospects.

The invention discloses a method and device for processing inter-cell interference, a control device and a base station. The method comprises the steps of: managing a first cell by a first base station, obtaining resource scheduling information of neighbor cells of the first cell in a first time interval, and determining the level of the inter-cell interference of the first cell according to the obtained resource scheduling information; and measuring the inter-cell interference value of the first cell in a second time interval, using the interference level determined according to the resource scheduling information of the first time interval as an index, and storing the inter-cell interference value of the first cell in the second time interval. Based on an association relation between the stored interference level and the practical inter-cell interference value obtained by measurement, the first base station can rapidly and accurately predict the inter-cell interference value in the future by determining the interference level of the inter-cell interference, and the influences of the fluctuation of the inter-cell interference on the accuracy of SINR estimation are further reduced.

The invention provides a method for determining nickel-based single crystal high-temperature alloy residual segregation. By means of a manufactured metallographic specimen, primary dendrite distance of nickel-based single crystal high-temperature alloy is measured, and mass fraction distribution of all elements on all points of the nickel-based single crystal high-temperature alloy are measured on the manufactured metallographic specimen through a dot matrix method. The segregation directions of the elements are determined according to obtained mass fractions of the elements on the measurement points. A computer simulates solution treatment to obtain distribution of the elements after solution treatment, and accordingly the residual segregation ratios of Al, Ta, W and Re elements are obtained respectively. According to the method, experiments and simulation are combined so that residual segregation of the high-temperature alloy can be accurately and quickly predicted after solution treatment, cost of solution treatment research is reduced, and the period of solution treatment research is shortened. The mode that the mass fraction distribution of the elements on the points is obtained through the dot matrix method has the advantage of being large in specimen number, and result deviation caused by large errors in common representing methods is prevented.

The invention discloses a voice interaction method, a model training method, electronic equipment and a storage medium. The voice interaction method comprises the following steps: acquiring user voice data to perform voice recognition in real time to obtain a user voice request; under the condition that the complete user voice request is not received, predicting the user voice request according to the user voice request acquired in real time and the recurrent neural network model to obtain a prediction result; processing the prediction result to obtain a first prediction instruction; and after the complete user voice request is received, if the prediction result is the same as the received complete user voice request, completing voice interaction according to the first prediction instruction. According to the method, the voice request of the user is predicted and complemented based on the recurrent neural network model, the total time required by a dialogue system for processing the voice request of the user is shortened, meanwhile, a lightweight model is adopted, conventional statements can be accurately and quickly predicted, and richness, parameter quantity, time cost and cost can be controlled.

The embodiment of the invention provides a financial service risk prediction method and device. The method comprises the steps of receiving a financial service risk prediction request for a target user; selecting one of the plurality of financial service risk prediction models as a target financial service risk prediction model based on the financial service request type, wherein the financial service risk prediction model is obtained by training by using respective corresponding training sets, and the training sets are target data sets obtained by performing data binning on historical user financial information of multiple users by using a Spark system; and inputting the user financial information into the target financial service risk prediction model, and taking output as a financial service risk prediction result of the target user. The reliability, efficiency and automation degree of the binning process of the training data can be effectively improved, and then the accuracy, efficiency and automation degree of financial service risk prediction of financial institution users by applying the financial service risk prediction model obtained by binning data training can be effectively improved.

The invention discloses an ANN-based radio frequency power amplifier interconnection reliability modeling method. According to the method, FEA calculation is adopted to generate training data and testdata; the training data and the test data are used for training and testing of an ANN respectively, and an ANN model is established; the ANN model is utilized to generate radio frequency power amplifier interconnection reliability data and establish a radio frequency power amplifier interconnection reliability database; and the radio frequency power amplifier interconnection reliability databaseis utilized to analyze interconnection reliability of radio frequency power amplifiers and influences of various factors on the interconnection reliability of the radio frequency power amplifiers. Through the method, the ANN is introduced into a radio frequency power amplifier interconnection reliability model research process, and before ANN model establishment is fully utilized, empirical knowledge relevant to a mechanism of a modeled object and an input / output relation is not needed, additional definition is not needed either, and therefore radio frequency power amplifier interconnection reliability modeling analysis efficiency is improved.

The embodiment of the invention discloses a resume filling method and device, electronic equipment and a storage medium. The method comprises the steps that resume information of a user is predicted based on user basic data and historical behavior data; And filling a resume for the user based on the predicted resume information. According to the scheme provided by the embodiment of the invention,the purposes of accurately and quickly predicting and filling the resume information of the user are achieved, and the problems that in the resume filling process in the prior art, the prediction dimension of the resume information is single, few predictable contents exist, the prediction result precision is poor, and the resume filling efficiency of the user is low are solved.

The invention discloses a rainfall prediction method and system based on wavelet decomposition. The rainfall prediction method comprises the following steps: acquiring a plurality of historical rainfall time sequences and a current rainfall time sequence; screening out the historical rainfall time sequences from the plurality of historical rainfall time sequences, wherein the relevancy between thehistorical rainfall time sequences and the currently-performed rainfall time sequence is greater than a set threshold value; decomposing the plurality of historical rainfall time sequences of which the relevancy is greater than a set threshold value and the currently performed rainfall time sequence to obtain a low-frequency component and a high-frequency component respectively; constructing a VAR model for the low-frequency component for prediction to obtain a low-frequency component prediction quantity; constructing an NAR neural network for the high-frequency component for prediction to obtain a high-frequency component prediction quantity; and reconstructing the low-frequency component predictor and the high-frequency component predictor to obtain a rainfall predictor.

The invention discloses a power line dangerous point prediction method and device based on unmanned aerial vehicle laser radar detection, and the method comprises the steps: carrying out the feature point segmentation of power grid environment point cloud data collected by a laser radar carried by an unmanned aerial vehicle, carrying out the fitting of the power line point cloud data, and obtaining the power line position information; and processing the obstacle candidate point data by using a surface construction rule to obtain surface position information of the obstacle candidate point, and obtaining a prediction result of the power line dangerous point by using the distance between the power line position information and the surface position information of the obstacle candidate point. The device comprises a feature point segmentation module, a position information construction module and a prediction module. According to the method and the device, on the basis of segmentation of ground feature points of an electric tower, a power line, low vegetation, ground, buildings and the like, curve fitting can be performed on the power line, and tall buildings and crown contours can be detected, so that quick and accurate prediction of dangerous points in the power line inspection process can be realized.

The invention discloses a MoRFs prediction model generation method and device, equipment and a storage medium. The method comprises the steps: acquiring a plurality of MoRFs segments and a plurality of non-MoRFs segments, wherein each MoRFs segment consists of a plurality of first sites, and each non-MoRFs segment comprises a plurality of second sites; extracting a first feature vector corresponding to each first site and a second feature vector corresponding to each second site; and training a pre-constructed initial prediction model by using the first feature vectors and the second feature vectors to generate a target prediction model which is used for predicting whether a site in a protein belongs to the MoRFs fragments or not. In this way, by means of the target prediction model, sites, belonging to the MoRFs fragments, in proteins can be conveniently, rapidly and accurately predicted.

The invention relates to an electricity selling company evaluation emotion classification method based on a joint attention mechanism, and belongs to the technical field of information. The method comprises the four steps of construction and division of an electricity selling company evaluation text data set, model training, model tuning and electricity selling company evaluation sentiment classification. The emotion evaluation classification result obtained by the method can reflect the emotional tendency of the power consumer to the power selling company to a certain extent, and can describe the customer satisfaction and favorable comment of the power selling company in the retail market in a certain sense; the classification result can be used as an important index for evaluating the service quality and credit rating of the electricity selling company, the electricity selling company can be promoted to continuously improve the service quality, optimize the electric power package structure and adjust the electric power price, and benign competition and healthy and stable operation of the electric power retail market are guaranteed.

The present application provides a hard disk failure prediction method and device, which relate to the technical field of computers. The method includes: reading the SMART attribute data of the hard disk to be detected according to a first preset time interval, wherein the SMART attribute data includes at least An attribute parameter; determine whether the SMART attribute data meets the negative sample write condition based on the attribute parameter, wherein the negative sample is used to indicate that the hard disk to be detected is in an abnormal operating state; when it is determined that the negative sample write condition is satisfied, the The SMART attribute data is written into the negative sample array to predict the fault status of the hard disk to be detected through the negative sample array. This application alleviates the technical problem of low prediction accuracy when using the prior art to predict the hard disk fault.

The present invention relates to an inspecting board for simulating bound condition of a first device and a second device, and utilizing said bound condition to judge that the pasty medium coated quantity and distribution range on interface between said first device and second device after which are bound are proper or not. Said inspecting board includes a simulation portion and sighting surface whose back is opposite to said simulation portion. After the inspecting board is combined with said second device, the simulation portion can be sighted by means of sighting surface, so that the diffusion condition of the pasty medium can be sighted so as to further obtain a standard coating quantity.

The present invention provides a method and device for predicting the Mach number of a wind tunnel based on a convolutional neural network. The method includes: acquiring wind tunnel data within a first preset time; preprocessing the wind tunnel data to convert the wind The tunnel data is converted into three-dimensional data; the convolutional neural network is used to train based on the processed wind tunnel data to obtain a prediction model; the wind tunnel data to be predicted is obtained, and the wind tunnel data to be predicted is input into the prediction model to obtain the wind tunnel data. The Mach number of the hole. The invention performs modeling according to the collected wind tunnel data, and can quickly and accurately predict the Mach number of the wind tunnel, thereby breaking away from the dependence on empirical formulas and improving the reliability and precision of Mach number measurement.

The invention discloses a fuel shaking noise simulation analysis method and platform and a computer storage medium. The method comprises the steps that a noise simulation model is established according to input fuel tank data of a vehicle; wherein the fuel tank data at least comprises bandage data, shock pad data and washboard data; fuel tank shell surface vibration data correspondingly output by the noise simulation model and node counterforce of the surface of a shock pad and a bandage mounting point are obtained according to input boundary conditions; wherein the boundary condition comprises an oil tank liquid level and a driving working condition; and based on a preset calculation rule, according to the surface vibration data of the fuel tank shell and the node counterforce of the surface of the shock pad and the bandage mounting point, a fuel shaking noise value is obtained. According to the fuel oil shake noise simulation analysis method and platform and the computer storage medium provided by the invention, the fuel oil shake noise can be quickly and accurately predicted under the condition that the whole vehicle data is not needed, the vehicle research and development efficiency is improved, and the vehicle research and development cost is reduced.