965results about How to "Achieve forecast" patented technology

The invention relates to the technical field of digital earth, in particular to a digital earth prototype system comprising a data receiving and fast processing sub-system, a metadata service sub-system, a model base sub-system, a space information database sub-system, a mesh calculation sub-system, a map application service sub-system and a virtual reality sub-system. With the adoption of the invention, it is able to construct a data resource sharing platform, a knowledge sharing platform, a computation resource sharing platform and a coordinated operation platform of the digital earth prototype system. Besides, with the establishment of the prototype system, the invention resolves relevant key technologies for the construction of a technology service platform for the application of space data and forms a service system for the application of space information in earth technology.

The invention provides a finite element prediction system for welding and solidifying a crack in virtual environment, which relates to the technical field of finite element prediction software for welding and solidifying a crack, and the complex simulation and the prediction processes of welding and solidifying a crack are simplified into the simple operation and settings of an interface. A preliminary processing subsystem comprises an interface operating module, an MSC.Marc interface document module, a subprogram.f file module and an invocation MSC.Marc software executing and computing module, and a post processing subsystem comprises a MSC.Marc software computed result processing module, a MatrixVB function module, a Matlab software module and a graphic display module; a solidified crack predicting subsystem comprises a resistivity curve loading module, a judgment module and a result output module. The application of the invention enables a common welding engineering technician to carry out the numerical simulation and prediction for welding and solidifying a crack; the computed result is processed, and the storage space of a result file is decreased; the result file is converted into a file format convenient for graphic display, and the computed results of a temperature field and a stress strain field are visually displayed in a graphic form.

The invention discloses a fault diagnosis prediction system of a large-span railroad bridge based on PHM. The system comprises a visualization management module, an archival data module, an online monitoring module, an artificial polling module, a diagnosis prediction module, a maintenance module, a data interaction module and a PHM database. The invention further provides a fault diagnosis prediction method of the large-span railroad bridge based on the PHM. The fault diagnosis prediction system has the advantages that 3S network architecture, a BIM and a GIS technology are adopted to associate and archive design, construction, operation and maintenance information, by comprehensively monitoring trains, tracks, bridges, and bridge environments and processing big data of artificial pollinginformation, the diagnosis and prediction of bridge diseases and evaluation of health states of the bridge are achieved, and a decision-making basis is provided for the maintenance of the bridges.

The invention discloses a federation learning method, a system and a readable storage medium. The federated learning method includes the following steps: the data terminal performs federation trainingon the multi-party training samples based on the gradient descent tree GBDT algorithm, to construct a gradient tree model, wherein the data terminal is a plurality of, the gradient tree model comprises a plurality of regression trees, the regression trees comprise a plurality of partition points, and the training sample comprises a plurality of features, the features correspond to the partition points one by one; the data terminal performs joint prediction on a sample to be predicted based on the gradient tree model to determine a prediction value of the sample to be predicted. The inventioncarries out federation training on multi-party training samples through GBDT algorithm, realizes the establishment of gradient tree model, and is suitable for scenes with large data volume and can well meet the needs of realistic production environment through the gradient tree model. Forecast the sample to be forecasted jointly, and realize the forecast of the sample to be forecasted.

The invention discloses a device and a method for evaluating oil paper insulation aging and predicting the service life of a transformer, and belongs to the technical field of transformers of an electrical power system. The device comprises a long-term aging device, a polymerization degree test device, a tearing strength test device and the like. An accelerated aging test is performed in the long-term aging device at the temperature of 130 DEG C with an oil paper mass ratio of 20:1; furfuraldehyde content in oil and polymerization degree of a piece of insulation paper are measured; the insulation paper is subjected to a mechanical strength test; the concentration change of furfuraldehyde in oil during a process of insulation paper aging is analyzed; the paper and water content during a paperboard aging process are tested; and gas content in oil is tested. Most of transformers are subjected to insulation aging and quit operation primarily due to the fact that the mechanical strength of the insulation paper is lowered and cannot support sufficient insulation. The device and the method which are provided by the invention can evaluate the insulation state of the transformer and predict the residual operation life of the power transformer. Therefore, the examination and repair can be reasonably arranged according to the equipment aging condition, the optimization of equipment use ratio is facilitated; and the management level of power enterprises is improved.

The invention relates to an off-grid grid-connected hybrid photovoltaic power generation control system and an economical operation optimization method of the off-grid grid-connected hybrid photovoltaic power generation control system. The control system is composed of a photovoltaic array, a grid-connected inverter, a storage battery pack, a bi-directional inverter, a load, a power meter, a public grid, a demand response control system and a switch set. The photovoltaic array is connected into an alternating-current side through the grid-connected inverter, and the storage battery pack is connected into the alternating-current side through the bi-directional inverter. The grid-connected inverter is connected with the bi-directional inverter through a switch, and the public grid and the load are connected through the switch set (S1-S4). The whole off-grid grid-connected hybrid photovoltaic power generation control system is uniformly managed and controlled by the demand response control system. Through the turn-on and turn-off combination of the switch set, the system supports eight different operation modes to achieve the optimization of economic benefits and peak shifting of the grid, and the service life of the system is prolonged through the method of limiting the discharging depth and charging and discharging power of storage batteries.

The invention provides a prediction method of a road traffic flow based on a graph convolution network. The method comprises the following steps of collecting vehicle GPS data in the past period of time; integrating the vehicle GPS data with actual road network information data to obtain a road traffic flow characteristic matrix, and converting a road and a road intersection point in the actual road network through line graph conversion to generate a road adjacency matrix; and based on the road adjacency matrix and the road traffic flow characteristic matrix, acquiring a predicted value of traffic flow data of the road in the next time period by using a spatial characteristic and a time characteristic of traffic flow data of the integrated road of a GCN network and an LSTM network. Through integrally using the spatial characteristic of the road traffic flow data extracted by GCN and the time characteristic of the road traffic flow data extracted by LSTM, and combining a time period characteristic and a weather characteristic, the road traffic flow data is predicted, and a prediction effect is better than the prediction effect acquired through using only time characteristic or the spatial characteristic.

The invention discloses a graph neural network traffic flow prediction method and system based on an attention mechanism. The method comprises the steps of obtaining to-be-predicted urban traffic flowdata; constructing a road network map according to the road connection relationship; preprocessing the urban traffic flow data to be predicted; and inputting the road network map and the preprocessedresult into a pre-trained neural network based on an attention mechanism, and finally outputting a prediction result of the urban traffic flow. Roads and checkpoints are encoded according to the roadnetwork information, a road network graph structure is established according to the upstream and downstream relationship of the roads, vehicle passing data of the checkpoints is counted under different time dimensions nd summarized to form a road network traffic flow data table; a graph neural network formed by stacking multiple layers of attention modules is constructed, a time sequence attention mechanism and the graph attention network are used for modeling the traffic flow in the whole road network, and the future traffic flow condition of a specified checkpoint is predicted.

The invention discloses a road network state prediction method based on a depth space-time convolution circulation network. The method comprises steps of step 1, establishing a sample set and dividing a training set and a test set; step 2, modeling a road network state prediction model, wherein the road network state prediction module comprises a convolution neural network and a recurrent neural network, the convolution neural network is used for extracting space features of the road network and the recurrent neural network is used for extracting sequential rules of the evolution of the road network; and step 3, carrying out road network state prediction in a certain period in the future. According to the invention, through the depth convolution neural network, relations between sections in the road network are extracted; through the recurrent neural network, the sequential rules of the revolution of the road network are extracted; and quite precise prediction of road network traffic states can be obtained by fully considering the space-time information of the road network.

The invention discloses a network security situation awareness model and method based on an attack graph. The model comprises a data preprocessing module, an attack feature library construction module, an attack state identification module, a security situation evaluation module and a security situation prediction module. The method comprises: firstly, collecting data sets of different sources, and obtaining asset vulnerability threat data and asset attack threat data after preprocessing; further analyzing and training by utilizing the asset attack threat data and the principal component dataof the attack information, and establishing an attack feature library; matching the attack portrait, and determining the attack state of the current equipment; then, calculating a security situation value of the whole network by utilizing the asset attack threat data; and finally, predicting the next attack state through the attack and defense game matrix of the attack state diagram so as to predict the network security situation. A good situation awareness effect and a network security situation evaluation and prediction function are realized, and the model and the method have relatively goodpracticability, high efficiency and expansibility.

The invention relates to the network safety, and provides an enterprise network safety event management system comprising safety devices, a safety event data acquisition module, a safety event data standardization module, a safety event situation assessment and prediction module, a safety device management linkage defense strategy generation module, a state protection module and a system management module. The method comprises the following steps: the system automatically collects safety event information generated by the safety devices in the enterprise network; the system carries out standardization format transition for the collected safety event information; the system analyzes the safety event information after standardization format transition so as to obtain network information safety situation, and predicts network safety trends; the system carries out association analysis for safety devices disposition safety application strategy according to the network information safety situation and the predicted network safety trends, thus generating the linkage defense strategy and sending the strategy to the safety devices. The enterprise network safety event management system and method are suitable for enterprise network safety event management.

The invention provides a transfusion monitor. The transfusion monitor comprises a shell, a fluid amount collecting and controlling device, a WIFI (Wireless Fidelity) communication module and a power supply management module, wherein the fluid amount collecting and controlling device, the WIFI communication module and the power supply management module are respectively arranged on the shell. The fluid amount collecting and controlling device is used for collecting relevant parameters in a transfusion device, and obtaining the dripping speed and residual liquid measure by the relevant parameters so as to obtain the residual time of the transfusion, realize the prediction on the transfusion completing time and achieve an early warning aim. The invention further provides a transfusion monitoring system based on a cloud service which comprises a transfusion monitor, a wireless receiving and dispatching module, a nurse station computer group, a cloud server and a patient terminal, wherein the transfusion monitor is used for transmitting monitoring information collected by the transfusion motor to the nurse station computer group through the wireless receiving and dispatching module, the nurse station computer group is used for transmitting the monitoring information to the cloud server, and the patient terminal and the nurse station computer group can visit the cloud server to obtain the monitoring information.

The invention discloses a sample prediction method based on federation training, comprises the following steps: federating two aligned training samples by using an XGboost algorithm to construct a gradient lifting tree model, wherein the gradient lifting tree model comprises a plurality of regression trees, and a split node of the regression tree corresponds to a feature of the training sample; based on the gradient lifting tree model, jointly predicting the sample to be predicted to determine a sample class of the sample to be predicted or to obtain a prediction score of the sample to be predicted. The invention also discloses a sample prediction device based on federation training and a computer-readable storage medium. The invention realizes federated training modeling by using trainingsamples of different data parties, and further realizes sample prediction based on the established model.

The invention provides a near-bit geology guide detecting system, comprising an electricalresistivity detecting unit, a gamma detecting unit, a position sensing unit and an insulating dipole transmitting unit, wherein the electricresistivity detecting unit is used for measuring the electricresistivity a near-bit stratum to generate electricresistivity data; the gamma detecting unit is used for measuring near-bit gamma rays to generate gamma ray data; the position sensing unit is used for measuring near-bit inclination and an implemental surface to generate position data; and the insulating dipole transmitting unit is used for transmitting the electricresistivity data, the gamma ray data and the position data to the ground through a wireless electromagnetic signal channel. A sensor is close to the near bit as much as possible, so when a drill drills into the sensor, the sensor makes an early prediction to information of the front part of the drill and information of the periphery stratum of the drill, thereby the aim of geology guide is achieved during drilling process.

The invention relates to a laminated and partitioned automatic current control method applicable to a multi-target power grid, and belongs to the technical field of automation of power system dispatching. The method comprises multi-region control, multi-target control and laminated and partitioned coordinated control of section current; a laminated and partitioned power grid current coordinated control system and an adaptive control strategy; a sorting and allocating strategy which is used for adjusting the output of machine sets in real time, comprehensively takes voltage class, power grid region and energy-saving dispatching principle into account, and comprises a sorting and allocating strategy of generator unit clusters of different voltage classes in different power grid regions and a sorting and allocating strategy of generator units of same voltage class in the same power grid region; and partitioned independent super-short-term load forecasting control carried out in each control region of the power grid. The laminated and partitioned automatic current control method disclosed by the invention can be flexibly applicable to various power grid operating modes, and ensures that the power grid runs stably in an important power supply section circuit overhauling mode; and a hydro-power generating unit and a thermal power generating unit are controlled to generate power more consistently, and the method provides a powerful technical support for energy-saving power dispatching.

The invention provides a dynamic people flow early warning method based on mobile network signaling data. The method includes the steps of: (1) separating a people flow early warning target region and people flow early warning observation regions; (2) counting and obtaining the people flow transition probability of each observation region distributed along with the travel time; (3) simultaneously counting a people flow leaving ratio in each period at the target region; and (4) predicting the total customer number from each observation region into the target region at the statistics time according to the real-time people flow in the observation regions and the people flow transition probability distributed along with the travel time, and predicting the customer number left from the target region before the statistics time according to the real-time people flow in the target region and the people flow leaving probability in the period at the region, thereby predicting the people flow size at the statistics time, and judging early warning grades. According to the invention, the current mobile network information is fully used, the target region and the observation regions are introduced, the statistics of the people flow at the target region is realized according to the people flow conditions of the target region and the observation regions, and the people flow early warning at the target region is finished according to a designate standard of the early warning grades.

The invention discloses a method and a device for predicting cell interference. The method comprises the following steps: dividing a coverage area of a cell to be predicted into grids, and determining latitude and longitude information corresponding to each grid; acquiring latitude and longitude information of a base station in the cell to be predicted, and acquiring latitude and longitude information and signal transmission power of a base station in a cell adjacent to the cell to be predicted; calculating a distance from each grid to each base station according to the latitude and longitude information of each grid and the acquired latitude and longitude information of the base stations; and determining interference values of the cell adjacent to the cell to be predicted for each grid according to a propagation model, the distance from each grid to each base station and the signal transmission power of the base station. The method and the device can realize the prediction for the cell interference.

A method for monitoring fatigue strength of driver includes shining driver eye with infrared ray to obtain multiimage of different retina image at the same time, carrying out difference processing for collected images to obtain pupil images, using kalmen filter to trace pupil in real - time to obtain pupil characteristic parameter, processing the parameter to obtain maximum value of pupil size and real - time coroclisis percentage of pupil, calculating out PERCLOS value f and using BP network sorter to judge fatigue strength of driver based on obtained value f. The monitoring device is also disclosed.

The invention discloses a regional mountain torrent risk prediction method and system. The method comprises the steps: S1, predicting the weather information of all hills in a region based on a meteorological model; S2, screening out hills needing risk prediction based on the weather information; S3, dividing hills needing risk prediction into different risk prediction levels according to the basic information of the hills; wherein different risk prediction levels correspond to different risk prediction periods; S4, preliminarily determining disaster-causing factors for mountain torrent evaluation; S5, screening the preliminarily determined disaster-causing factors to obtain main disaster-causing factors influencing the mountain torrent; S6, training and generating a plurality of mountaintorrent risk prediction models based on the main disaster-causing factors; S7, several risk prediction models with the best performance are selected to be combined to form a final risk prediction model; and S8, performing mountain torrent risk prediction on the hill to be predicted. According to the method, the risk prediction of the regional hill is realized, the realization cost is low, the coverage is wide, the processing efficiency is high, and the safety of the hill is improved.

The present invention relates to a rolling bearing fault diagnosis method and system based on a convolutional neural network. The method comprises the following steps: the step S1, constructing a database configured to reflect the work condition of a rolling bearing; and the step 2, identifying the work condition of the current rolling bearing through the database. The rolling bearing fault diagnosis method and system based on the convolutional neural network obtain the vibration signals of the rolling bearing through an accelerometer, and the fault diagnosis method of the rolling bearing employs the powerful image identification capability of the convolutional neural network model to identify faults of the rolling bearing through deep learning of big data so as to realize the monitoring and forecasting of the faults of the rolling bearing and substantially improve the accuracy and the timeliness of the rolling bearing fault diagnosis.

The invention discloses a method and a system of moving target tracking based on a video. The method of moving target tracking comprises the following steps of: detecting a moving target, and fragmenting the moving target in a video image; establishing a color probability distribution map for the moving target; calculating a barycenter of a search window, and adjusting the size of the search window according to window zeroth moment; predicating the position of the search window in the next frame of image through a Kalman filter; and carrying out barycenter matching in an estimation range to track the target. According to the method and the system for moving target tracking based on the video, the problem of moving target tracking at complicated background can be well solved, and the method and the system are good in real-time performance and robustness.

The invention provides a simulation method for residual stress prediction of micro-milling nickel-based superalloy, belongs to the field of finite element simulation cutting, and relates to a finite element simulation method for residual stress prediction analysis in the cutting processing process of difficult-to-process materials of the nickel-based superalloy. According to the simulation method, the finite element simulation method is used; the elastic-plastic constitutive relationship of materials is considered; and a Johnson-Cook fracture failure criterion is used for judging the failure, so that residual stress prediction values under different cutting parameters can be obtained. A micro-milling cutter for experiments is shot into a picture through a scanning electron microscope; the picture is drawn into a micro-milling cutter solid model through software; a three-dimensional processed workpiece model is built; and the model is subjected to lattice division. The method has the advantages that the surface residual stress can be measured, and the residual stress in the depth direction can also be measured, so that the problem of measurement difficulty of the residual stress in the depth direction in the measurement process is solved. The cost is reduced, and the labor is saved. In a prediction model, the cutting parameters are variable; and the prediction on the surface residual stress of the difficult-to-process materials is effectively realized.

The invention discloses a vision auxiliary device and method for automatic landing of an unmanned aerial vehicle, and relates to the field of unmanned aerial vehicles. The problems that according to an existing vision auxiliary device which acquires and recognizes runways and the horizon through an airborne photographing device so as to sense the landing posture of the unmanned aerial vehicle, due to the lack of a multi-scale analysis mechanism, the landing requirement cannot be easily met, the posture and position calibration effect is poor, and deep prediction of the movement behavior of the unmanned aerial vehicle lacks are solved. The vision auxiliary device comprises a two-shaft rotary table, a rotary table motor, a power module, a single-view fixed-focus camera, a vision auxiliary information processing module and an information storage module. The vision auxiliary information processing module comprises a landing track re-planning module, a data comparator, an information filter, a computer vision processor based on image control, a computer vision processor based on position control, a first feedback correction device, a second feedback correction device, an information fusion parameter extractor and an RS485 communication interface. The vision auxiliary device is simple in structure, low in power consumption, high in efficiency and good in compatibility.

The invention provides a complete-period capacity prediction method and device of tight oil multiple-media coupling seepage. The method comprises the following steps: obtaining a model parameter of an oil field development area and a multiple-media seepage mechanism parameter in the oil field development area; on the basis of a seepage mechanism in the oil field development area, the non-instantaneous transmissibility of pressure and the regional performance of seepage, establishing a capacity prediction model of a production well in the oil field development area; and according to the model parameter and the seepage mechanism parameter, utilizing the capacity prediction model to carry out iterative computation to obtain a production dynamic curve of at least one production stage of the production well so as to carry out the capacity prediction of tight oil. The method and the device consider influences including the complex seepage mechanism, the regional performance of seepage and the like, and can obtain a more accurate capacity prediction result.

The invention discloses a social network link prediction method based on multi-source heterogeneous data fusion. Link prediction is carried out by using a social network based on geographic position information, wherein the social network comprises two heterogeneous data sources, namely a user relationship topological graph and a user sign-in record. The invention provides a mixed frame. The association between two heterogeneous data sources, namely a user relationship topological graph and a user sign-in record, in a social network based on geographic position information is fully captured through a model AL. The problem that the prediction result is inaccurate when a single data source in a social network based on geographic position information is used for link prediction is solved, andthe link prediction effect is effectively improved. And meanwhile, local sensitive hashing is applied to improve the calculation speed of deep learning for training and reduce the storage overhead.

The invention relates to the field of bioinformatics and data mining, in particular to a hierarchical multi-label classification method for protein function prediction, and aims at solving the data set imbalance problem, multi-label problem and hierarchical constraint problem when the conventional classification methods are used for predicting protein functions. The method comprises the following steps of: 1, a training stage: training a data set of each node in a class label hierarchical structure by adopting an SVM classifier in the training stage so as to obtain a group of basic classifiers; and 2, a prediction stage: firstly obtaining preliminary results of unknown samples by using the group of basic classifiers obtained in the training stage in the prediction stage, and processing the results by adopting a TPR algorithm with a weight so as to obtain a final result which satisfies a hierarchical constraint condition and realize the prediction of the protein functions. The hierarchical multi-label classification method for protein function prediction is applied to the field of bioinformatics and data mining.

The invention discloses a method and a device for predicting disease, an equipment and a computer-readable storage medium. The disease predicting method comprises acquiring a medical image in responseto a disease prediction instruction; detecting the medical image to obtain position information of at least one lesion; extracting a lesion image for each of the lesions from the medical image basedon the position information. A lesion image of the lesion is described; each of the lesion images is input to a disease prediction model to obtain a corresponding disease prediction result. By adopting the invention, the prediction of the disease can be realized, so that the diagnosis efficiency and the accuracy of the disease can be improved.

The invention discloses a star sensor star point image enhancement method, which is a superposition enhancement method after reconstruction of a star point image. Through motion information acquired by the star sensor and a star point extraction and recognition result of the current frame, a nonlinear model for star point motion is adopted to realize star point position prediction and tracking andrecognition in a next frame, and position prediction results of star points at different positions in an image plane are more accurate; by using the predicted star point position and the star point information in the former frame, a star point motion energy distribution model is used to generate one frame of noise-free simulated star map, the energy distribution is closer to the actual star points, no complex matching alignment process is needed, better star point superposition enhancement effects can be obtained, and the star point extraction precision is improved. The method is suitable when the star sensor is in a static condition and a dynamic condition, star point position prediction and tracking and recognition can also be realized, the star sensor processing time is further reduced, and the star sensor attitude output errors are reduced.

The invention discloses a method and system for predicting an insulation resistor fault of an electric car, and the method comprises the steps: building the system through employing a synchronous voltage sampling circuit, a control unit, a whole vehicle controller, a first bias resistor, a second bias resistor, an electronic switch K1, and an electronic switch K2; measuring a nominal voltage of a serial battery pack of the electric car and the nominal voltages of all cell modules; and calculating the equivalent parallel insulation resistance and the equivalent voltage source voltage of the system. The resistance values of all cell modules in the serial battery pack of the electric car are compared with the equivalent parallel insulation resistance, thereby achieving the on-line diagnosis of an insulation fault and the on-line positioning of the insulation fault. A plurality of equivalent parallel insulation resistance values of the system at different moments are calculated, and a curvilinear equation of time change with the equivalent parallel insulation resistance values is fitted, thereby achieving the prediction of occurrence remaining time of insulation faults of all stages.

The invention discloses a short-term wind speed multi-step prediction method based on a deep learning method, and relates to a short-term wind speed multi-step prediction method, which aims at solving the problem of poor wind speed prediction effect in an existing wind speed prediction method. The method comprises the following steps of 1, on the basis of the deep learning method, building a deep neural network regression model with a multiple input multiple output structure; 2. using of a gradual-layer greedy method, combining the recent actual wind speed data of a measured wind farm to train the built deep neural network regression model, and learning and obtaining the mapping relationship among the sequences of the obtained model through the non-linear mapping function of the model to determine the deep neural network regression model; and 3. according to the determined deep neural network regression model, carrying out multi-step prediction on the actual wind speed of the measured wind farm, so as to obtain the wind speed prediction result of the measured wind farm. The method is used for predicting the short-term wind speed of the wind farm.