29results about How to "Achieving Precise Forecasting" patented technology

The invention provides a new energy storage consumption capability index prediction method, and relates to the field of power system operation, and the method comprises the steps: building a nonlinear mapping relation between solar radiation intensity, photovoltaic module conversion efficiency, a temperature coefficient, an aging attenuation rate factor and power generation efficiency for photovoltaic power generation, so as to dynamically correct a photovoltaic power generation potential output prediction value; based on a potential output prediction result of photovoltaic and wind power generation, solving an energy storage capacity configuration and operation strategy through an intelligent optimization algorithm to maximize the benefit of an energy storage system, and obtaining a dynamic prediction value of an energy storage consumption capability index; and combining prediction results of the statistical model based on the historical data, the dynamic model based on the physical mechanism and the data-driven model based on machine learning by adopting an adaptive weighted fusion strategy. According to the method, accurate prediction of potential output of new energy power generation and dynamic evaluation of the energy storage absorption capability are realized, and effective technical support is provided for new energy grid connection and optimal configuration of an energy storage system.

The invention relates to a radio frequency module electromagnetic compatibility analysis method and system, and the method comprises the steps: obtaining the structure data and electromagnetic parameter information of a radio frequency module, carrying out the electromagnetic simulation of the structure data based on the electromagnetic parameter information, and obtaining the corresponding simulation data of the radio frequency module; acquiring near-field scanning data of the radio frequency module, and performing correlation calculation on the near-field scanning data and the radio frequency module simulation data through a preset fully implicit frequency domain finite volume numerical algorithm to obtain corresponding far-field electromagnetic distribution characteristic data; obtaining interference emission data of the radio frequency module, and performing zero eigenvalue entropy correction to obtain corresponding interference propagation feature data; performing optimization analysis on the far-field electromagnetic distribution characteristic data and the interference propagation characteristic data to obtain corresponding initial optimization information; and performing scheme analysis on the initial optimization information and an EMC performance evaluation result to obtain a corresponding EMC optimization scheme. According to the invention, the electromagnetic characteristics of the radio frequency module can be evaluated more comprehensively, so that the accuracy of electromagnetic compatibility analysis is improved.

The invention belongs to the technical field of machine learning, and discloses an earth and rockfill dam burst flow prediction method based on data enhancement and an adaptive network. Comprising the following steps that historical earth and rockfill dam outburst flow data are collected and preprocessed, and then historical earth and rockfill dam outburst flow data are obtained; designing and training a generative adversarial network, and expanding the historical earth and rockfill dam outburst flow data through the generative adversarial network to obtain an earth and rockfill dam outburst flow data set; performing feature extraction and weighting on the data in the data set to obtain outburst flow feature data; constructing a support vector machine model, and training the support vector machine model by using an adaptive particle swarm optimization algorithm to obtain a trained support vector machine model; according to the method, the training set is enhanced by using the generative adversarial network model and the attention mechanism, and the enhanced training set is input into the adaptive particle swarm optimization support vector machine model, so that accurate prediction of the earth and rockfill dam burst peak flow can be realized.

The invention belongs to the technical field of sea wave height prediction, and discloses a sea wave height prediction method and system based on a large language model and a medium, and the method comprises the steps: S1, geographic position coding: carrying out geographic position coding on sea wave height data, and generating sea wave height data fused with geographic position information; s2, a step of ocean wave height semantic feature coding based on an attention mechanism: carrying out self-coding on the ocean wave height data output in the step S1, outputting ocean wave self-coding features, carrying out self-coding on ocean wave physical field features, outputting word vector features, and realizing interaction between the ocean wave self-coding features and the word vector features by utilizing the cross-modal attention mechanism; outputting domain-specific semantic features; and S3, a step of large-scale language model prediction based on sea wave height domain cue word driving: inputting the domain specific semantic features into a large-scale language model, and outputting prediction features. According to the method, sea wave complex spatial-temporal change information is effectively mined, and accurate prediction of the sea wave height is realized.

The invention relates to an efficient deployment method and system for a privatized large language model ecological service. The system comprises an evaluation module, an index collector, a monitoring module, a vLLM framework, a model cluster service module and a request distribution module. The evaluation module performs quantitative evaluation on the performance of the system under different hardware platforms; the index collector collects server cluster performance indexes, cluster resource states and load information in real time; the monitoring module is used for monitoring the server cluster; the vLLM framework dynamically allocates computing resources to realize management of a multi-modal server cluster, and performs unified management and scheduling through PM2 to realize multi-model efficient parallel task processing; a vLLM framework is started by the model cluster service module, and different large model instances are packaged in a Docker container; the request distribution module distributes external requests in a self-adaptive mode according to cluster resource states and load information fed back by the index collector in real time. Intelligent request distribution and dynamic load balancing are realized, and integration and efficient management of a multi-modal large language model are realized.

The invention discloses an LED display screen field correction method and a related device, and relates to the technical field of display, and the method comprises the steps: carrying out the preliminary detection of a plurality of LED units, and recording the initial display parameters of the plurality of LED units; determining a target display parameter of the whole screen; adjusting and calculating the plurality of LED units to obtain a parameter adjusting scheme of the plurality of LED units; collecting life cycle data of the plurality of LED units and analyzing the life cycle data, and respectively determining predicted residual life of the plurality of LED units under the parameter adjustment scheme; screening out a plurality of target LED units of which the predicted residual life is within a preset durability range; and performing point-by-point correction on the target LED unit based on the parameter adjustment scheme. By implementing the method, residual life prediction and accurate correction are performed on different batches of LED units, so that reasonable utilization of the tail LEDs is realized, a display screen shows better performance uniformity and longer service life, and user experience is improved.

The invention discloses a power distribution network fault early warning method considering operation state uncertainty, and the method comprises the following steps: calculating each operation parameter of a power distribution system based on the historical data and prediction data of distributed photovoltaic and load, and calculating the difference value of the operation parameters under the two conditions; determining a time-phased probability distribution model of the parameter difference, and generating an operation parameter set based on the probability distribution model; according to the calculated parameters, constructing a future operation parameter prediction model based on LSTM, superposing the operation parameter set to obtain a prediction result of the operation situation of the power distribution network, and performing fault early warning on the power distribution network according to the superposition result; according to the method, the future operation situation of the power distribution network can be sensed and pre-judged in advance, and a powerful guarantee is provided for safe and stable operation of a power system.

The invention discloses a method for positioning and predicting a wireless signal equal-strength boundary value based on a motion vector set. The method comprises the following steps: constructing an offline map of an indoor scene in a server; performing grid division on the scene, acquiring a signal intensity value of the Bluetooth beacon at each grid center point, performing intensity value region division on the signal intensity values, and performing numerical fitting to obtain an intensity boundary line; when a user carries wireless equipment to move in a scene, cutting an intensity boundary line to obtain a dynamically updated movement vector set; at each moment, a position estimation area is constructed according to the position estimation results of the multiple motion vectors; and fusing a plurality of position estimation results through Bayesian estimation, namely predicting the position at the current moment in combination with historical data before the current moment. According to the invention, a strategy of first mapping and then positioning is adopted, a user downloads a scene map through a server, information fusion calculation is realized through a positioning algorithm, and accurate positioning and prediction of the user are realized.

The invention discloses a ship track prediction method based on Attention-BiLSTM, and relates to the technical field of ship track prediction. The method comprises the following steps: acquiring ship AIS data, weather data and date attribute data; carrying out abnormal data detection and elimination on the ship track data, and filling missing data after elimination; establishing a ship track feature input candidate set, selecting a feature input variable influencing the ship track, determining a delay order between the ship track and the feature input variable, and constructing a feature vector; dividing the feature vector into a limited number of sub-model action ranges, and performing boundary division and boundary estimation; and establishing a ship track prediction model for each sub-model action range, and realizing rolling prediction of the ship track through the ship track prediction model. Accurate high-quality data and reasonable action range division are provided for ship track prediction, the ship track prediction precision is guaranteed, and the purposes of ship reasonable path planning, resource optimization scheduling and marine navigation safety improvement are achieved.

The invention belongs to the field of TBM shield tail gap measurement, and particularly relates to a method for non-contact measurement of a gap between an open TBM shield tail and a surrounding rock, and the method combines Blender software to establish a virtual scene comprising the surrounding rock, a TBM shield, a reflective sticker, a binocular camera and an LED spotlight. The textures of the TBM shield, the surrounding rock and the reflective sticker in the virtual scene can be set according to a real image, a large number of binocular camera images can be rendered through changes of the textures and the surrounding rock surface, meanwhile, a gap between the tail of the TBM shield and the surrounding rock under a corresponding binocular graph is obtained through a written blender script, and a data set is generated. The data set is used for building a neural network model on a server, training is carried out, and a training weight is obtained; and then deploying the model and the weight into a measuring device, namely shooting an image of the binocular camera through the measuring device, and measuring the gap between the tail part of the shield and the surrounding rock.

The invention discloses a method and device for predicting the number of leiocassis longirostris populations and a storage medium. The method comprises the steps that the number of fish schools is obtained, and the length and width of fishes, the scanning area, the detection distance and the detection depth are detected; inputting the length and the width of the detection fish into a pre-configured standard frequency estimation model to obtain an occurrence frequency estimation value; performing weighted calculation according to the occurrence frequency estimated value and the equivalent area to obtain an occurrence frequency weighted value; wherein the equivalent area is an area value which is calculated according to the length and width of the detected fish and is used for representing the size of the fish; performing product calculation according to the scanning area, a preset transverse rotation angle and a preset vertical rotation angle to obtain a detection water area; carrying out mean value calculation according to the detection depth to obtain a detection average depth; calculating an effective volume; and based on a preset multi-source data fusion model, performing calculation in combination with the effective volume, the fish school number, the occurrence frequency weighted value and the detection distance to obtain the population number. The method can improve the prediction accuracy.

The invention relates to a long-sequence prediction method for key production variables in a polyester fiber polymerization process, and the method comprises the steps: inputting the historical data of a certain time length of a plurality of key production variables in the polyester fiber polymerization process into a proposed time anchor network model, and outputting a predicted value of a future time step of the plurality of key production variables; an encoder of the time anchor point network model is a plurality of time anchor point layers, and each time anchor point layer is formed by connecting a time anchor point sampling module and a graph mask full-connection neural network in series; a decoder of the time anchor network model is a two-layer full-connection neural network. According to the method, modeling is carried out on a multivariable prediction task of an aggregation process from the time scale and spatial correlation, the proposed anchor point sampling-based technology contains the idea of division and treatment, data is subjected to block sampling, different neural networks are adopted to extract features, a time-based anchor point network model is established, and the time-based prediction task of the aggregation process is obtained. The problem of long sequence modeling can be effectively solved, and accurate prediction of key production variables is realized.

The invention provides an intelligent management system for the number of commodities in a shopping mall, and the system comprises the steps: 1, collecting key data of sold commodities, 2, carrying out the data modeling of the collected key data through a graph database, 3, analyzing the data modeling through a time sequence, 4, converting the analysis prediction result into a specific operation instruction, and 5, carrying out the management of the number of commodities in the shopping mall. The commodity management platform automatically updates commodity inventory data according to confirmation of the shop owner on the order information, the internal inventory state and the feedback of the supplier; key data of sold commodities are collected, data modeling is carried out on the key data, data modeling is analyzed through a time sequence, an analyzed prediction result is converted into a specific operation instruction, and the commodity management platform carries out commodity management according to confirmation of shop owners on order information and feedback of internal inventory states and suppliers. Commodity inventory data are automatically updated; and the real-time inventory and the predicted replenishment time are displayed, and the threshold value of the inventory is dynamically adjusted according to the market demand.

The invention discloses a water quality prediction method and system based on an artificial neural network, and relates to the technical field of water quality monitoring, and the method comprises the steps: obtaining a water quality data set, and carrying out the preprocessing of the water quality data set; randomly dividing the preprocessed water quality data set into a training set and a test set; constructing a multi-layer perceptron model, a long and short-term memory network model and a clustering multi-layer perceptron model; training the multilayer perceptron model, the long short-term memory network model and the clustering multilayer perceptron model by using the training set, and testing the trained multilayer perceptron model, the long short-term memory network model and the clustering multilayer perceptron model by using the test set; comparing root-mean-square errors of the multi-layer perceptron model, the long-short-term memory network model and the clustering multi-layer perceptron model on the test set, and selecting the model with the minimum root-mean-square error as the optimal model. According to the invention, through real-time online monitoring of water quality data, accurate prediction of the dissolved oxygen content is realized, so that the reliability of water quality monitoring is improved.

The invention discloses a defect severity prediction method based on developer modification history, and belongs to the technical field of software development. Comprising the steps of extracting modified historical data and defect report data, constructing an original data set, preprocessing the original data set, establishing a standard data set according to a file-defect-developer mapping relation, and constructing a developer-file weighted directed network based on the standard data set. Obtaining a file-file weighted directed network through weighted projection based on the developer-file weighted directed network, calculating a structure entropy score of a file node based on the file-file weighted directed network, and extracting a software measurement index of each source code file; and constructing a comprehensive feature set based on the structure entropy score, the software measurement index and the defect label combination, and training and verifying a machine learning model based on the comprehensive feature set to realize the defect severity prediction of the file. According to the invention, the accuracy and stability of defect severity prediction are improved from the perspective of history modification by developers.

The application provides a vibration environment prediction method for aviation equipment structure, relates to the field of environmental engineering, and comprises the following steps: S1, measuring vibration responses of known vibration points; S2, determining demarcation points of a path L1; S3, partitioning a region of a structure plane; S4, solving frequency response functions from the known vibration points to designated to-be-measured points in each partition; S5, solving vibration responses of the designated to-be-measured points in each partition; and S6, sequentially solving vibration responses of multiple designated to-be-measured points in each partition. The point partitioning method provided by the application adopts the similarity of the power spectrum density curves of the measured responses as a partitioning index, compared with the traditional partitioning method based on the physical structure, is more scientific and reasonable, and is beneficial to realizing accurate prediction of the vibration environment.

The invention discloses a multi-scale geological prediction method for risky sections of shafts and tunnels based on air-well-hole combination, which relates to the technical field of multi-scale geological prediction for risky sections of shafts and tunnels. A pre-mining geological survey is conducted on the surface of the area where two parallel mine tunnels are located, and the risky geological sections are delineated in combination with geological disaster characteristics; an aero-electromagnetic detection is conducted on the surface above the risky geological sections corresponding to the two parallel tunnels, and whether the geological body between the two tunnels has a water-containing body or a water-conducting structure is judged according to the detection result; according to the judgment result, an oblique drilling is conducted from the risk-free tunnel to the geological body behind the mining face of the high-risk tunnel, and the drilling parameters in the drilling process are obtained, and the drilling slag is collected; based on the drilling slag, the lithology of the rock mass, the type of unfavorable geological body and the degree of abnormality in the geological body between the two tunnels are judged in combination with spectral technology; an acoustic wave test is conducted in the hole, and the integrity of the rock mass is judged in combination with the drilling parameters and the acoustic wave velocity value, so as to complete the multi-scale accurate geological prediction for the risky sections of shafts and tunnels.

The application discloses a positioning and prediction method of radio signal equal-intensity boundary value based on a mobile vector set, comprising the following steps: constructing an offline map of an indoor scene in a server; dividing the scene into grids, collecting signal intensity values of Bluetooth beacons at the center of each grid, dividing the signal intensity values into intensity value regions, and obtaining intensity boundary lines after numerical fitting; a user carries a wireless device to move in the scene, cuts the intensity boundary lines, and obtains a dynamically updated mobile vector set; at each time, a position estimation region is constructed according to position estimation results of multiple mobile vectors; and multiple position estimation results are fused through Bayesian estimation, that is, the position at the current time is predicted by combining historical data before the current time. The application adopts a strategy of mapping first and then positioning, the user downloads a scene map through the server, and realizes information fusion calculation through a positioning algorithm, so that accurate positioning and prediction of the user are realized.

The invention belongs to the technical field of meteorological prediction, and discloses a typhoon intensity prediction method, system and device and a medium, so as to solve the problem of poor typhoon intensity prediction accuracy. The method comprises the following steps: acquiring typhoon initial field data, wherein the typhoon initial field data comprises at least two pieces of ERA5 reanalysis data at previous moments; a Qihu weather big model is obtained, typhoon initial field data are input to the Qihu weather big model, typhoon prediction initial data at the later moment are output, and the typhoon prediction initial data comprise the typhoon center position and typhoon characterization variables; calculating other characterization variables based on the output typhoon characterization variables to obtain channel dimension data; selecting channel dimension data in a preset diameter range by taking the center position of the typhoon as a reference to obtain spatial dimension data; and obtaining preset time sequence parameters, arranging the spatial dimension data according to a time sequence to form time dimension data, and inputting the time dimension data into the obtained systematic deviation correction model to output typhoon intensity prediction data at a future moment.

The invention discloses an early warning method and system for predicting soil slope damage based on shear parameters, and the method comprises the steps: obtaining monitoring data of a target region at a target moment, and predicting the moisture content of the target region in combination with a moisture content prediction model, the monitoring data comprising consolidation condition data monitored by a pressure sensor; predicting a target shear strength parameter of the target area at the target moment based on a shear parameter prediction model in combination with the predicted water content and the consolidation condition data; and determining an early warning level corresponding to the target area according to the target shear strength parameter, and performing early warning according to an early warning mode corresponding to the early warning level.

The application discloses a laser processing quality prediction method based on an improved starfish optimization algorithm and relates to the technical field of neural network models. The method comprises the following steps: first, improving the starfish optimization algorithm based on the Levy flight principle to obtain an improved starfish optimization algorithm; second, optimizing the weights and biases of a back propagation neural network based on the improved starfish algorithm, and inhibiting overfitting of the back propagation neural network based on an L1 regularization method to obtain a target back propagation neural network; third, acquiring process parameters of silicon carbide laser processing; and finally, inputting the process parameters into the pre-trained target back propagation neural network to obtain the processing quality parameters of the silicon carbide laser processing. The starfish optimization algorithm is improved by introducing the Levy flight principle, and high-precision prediction of the silicon carbide picosecond laser processing quality is realized.

The invention provides a digital pathological image grading method and system, and the method comprises the steps: carrying out the cell recognition based on a Unet algorithm which carries out the coding, decoding and regression to the classification of pixel points with the same size as an original image, and extracting a cell part in a prostate pathological image; inputting the extracted cell part into a pre-training model and a classification algorithm based on a convolutional neural network (CNN), and detecting cancer cells contained in the cell part; and based on the detected relevant features of the cancer cells and the surrounding tissues thereof, predicting the pathological grading of the cancer cells through an algorithm of a network model CNN-RNN combining a pre-training model, a convolutional neural network and a recurrent neural network. The technical problems that the workload is large, the interpretation threshold is high, and errors and subjectivity exist in judgment of the proportion of cancer cells by naked eyes are solved.

The invention relates to the technical field of intelligent traffic systems, and particularly provides a short-time traffic flow prediction method and system, and the method comprises the steps: obtaining a traffic flow sequence; decomposing and reconstructing the traffic flow sequence through an ensemble empirical mode decomposition method to obtain a plurality of traffic flow sequence components and residual components; calculating the correlation between the traffic flow sequence component and the traffic flow sequence according to a dynamic time warping algorithm to obtain a high correlation component and a low correlation component; performing calculation according to the high-correlation component and the low-correlation component to obtain normalized data; inputting the normalized data into an LSTM model to obtain a plurality of initial prediction results; and carrying out superposition summation on the plurality of initial prediction results to obtain a prediction result. The prediction accuracy is remarkably improved, the complex and changeable characteristics of the traffic flow can be flexibly handled, and a solid and reliable decision basis is provided for a traffic management department.

The invention relates to a multi-stage loading cumulative plastic strain prediction model based on MICP improved construction waste, a model construction method and a prediction method. On the basis of a test result of multi-stage loading cumulative plastic strain, a cumulative plastic strain prediction model which comprehensively considers factors such as calcium ion concentration, maintenance time, dynamic stress, matrix suction and loading times is established. According to the model, the strain evolution process of the construction waste under different working conditions is effectively described by introducing the coupling effect of multiple factors, so that the long-term deformation characteristic of the material is accurately predicted. A multi-stage loading cumulative plastic strain prediction model is established, and the method has important significance on performance evaluation and application optimization of engineering materials under different stress conditions. Through the prediction model, the mechanical performance of the construction waste can be evaluated in advance, reasonable selection of materials in different application scenes is guided, and long-term safety and stability of the structure are ensured.

The invention provides a dual-phase steel forming quality prediction and process adjustment method, which relates to the technical field of dual-phase steel forming, and comprises the following steps: establishing a multi-dimensional data acquisition network; obtaining a process parameter combination; performing mapping correlation analysis on dual-phase steel forming quality fluctuation in the temperature field analog simulation space and the stress field analog simulation space in combination with process parameter combination; dividing the process parameter combination; and by taking a preset forming quality standard as a constraint condition, performing learning training on the second type of technological parameter components, the third type of technological parameter components and the fourth type of technological parameter components, and performing technological parameter adjustment. The technical problem that in the prior art, due to the fact that interaction of multiple technological parameters in the dual-phase steel forming process is complex, forming quality fluctuation is difficult to control accurately can be solved, accurate prediction of the forming quality and intelligent adjustment of the technological parameters are achieved by establishing the multi-dimensional data collection network and the analog simulation space, and the forming quality is accurately predicted. And the stability of product quality is ensured.

The invention discloses an urban river water quality intelligent early warning system based on multi-parameter on-line monitoring, and relates to the field of water environment monitoring and control, and the system specifically comprises the steps: collecting water quality data in real time through arranging a multi-parameter sensor array, extracting spatial-temporal characteristics through employing a double-flow time sequence convolution network and an adaptive attention layer, and carrying out the real-time monitoring of the water quality. And performing anomaly evaluation based on the expert knowledge base. According to the system, a river branch water system is mapped into a time-varying directed graph structure for pollution tracing, gate linkage control is executed by adopting a distributed edge computing architecture, and pollution surrounding control and emergency disposal are achieved.

The application discloses a kind of intelligent computing center energy storage control method, device, system and storage medium, belong to intelligent computing center power supply and distribution technical field, the method includes: obtaining the multi-source data of intelligent computing center, including electrical quantity data and GPU task scheduling data;Multi-source data is input to first time series prediction model, obtain the load fluctuation prediction result of future millisecond level period, including predicted load fluctuation time, load fluctuation amplitude and fluctuation direction;Based on fluctuation amplitude and direction, reverse pre-charging and discharging operation is carried out to energy storage unit within the preset length of time before fluctuation occurs;After fluctuation occurs, feedback correction amount is generated according to the deviation of actual load power and predicted load value;Feedback correction amount and feedforward power control amount are superimposed to generate composite control amount, control the charging and discharging power of energy storage unit.The technical problem that related art is poor in fluctuation suppression effect due to the delay of response, lack of prediction ability is solved.

The present disclosure provides a method for predicting the prognosis of non-small cell lung cancer based on multimodal fusion, including obtaining a pre-treatment CT image and a radiotherapy three-dimensional dose distribution image of a target patient; obtaining a region of interest based on the pre-treatment CT image; obtaining a multi-omics prediction label based on the pre-treatment CT image, the radiotherapy three-dimensional dose distribution image and the region of interest; obtaining a deep survival label based on the pre-treatment CT image, the radiotherapy three-dimensional dose distribution image and the region of interest; and realizing the prediction of the patient's PFS based on the multi-omics prediction label, the deep survival label and the prognosis-related clinical features. The non-small cell lung cancer prognosis prediction method based on multimodal fusion disclosed in the present disclosure, combined with multimodal data such as CT images and dose distribution images, can fully combine the multimodal complementary feature information contained in the primary tumor and the affected lymph nodes, and can process the truncated data in the prognosis label, thereby realizing the accurate prediction of the patient's progression-free survival after radiotherapy.

The application relates to the technical field of weather radars, in particular to a radar intelligent echo extrapolation method based on a global-local aggregation model.A kind of radar intelligent echo extrapolation method based on global-local aggregation model is provided according to the application, a global-local aggregation model is built and trained by constructing meteorological radar echo grayscale image sequence dataset and its corresponding optical flow sequence dataset, the accurate prediction of future meteorological radar echo image sequence is realized.Carrying out the echo sequence optical flow information as motion guide information, and effectively fusing the echo sequence space-time information under different time scales by means of the attention mechanism, emphasizing the global-local space-time aggregation urgently needed in the echo extrapolation task, the ability to predict the long time sequence of strong rainfall coverage area and intensity is improved.