207results about How to "Easy to predict" patented technology

The invention relates to the technical field of autonomous robot decision-making systems, in particular to an autonomous robot decision-making system and method based on multi-modal sensing fusion, and the system comprises a multi-modal data collection module which is used for obtaining multi-modal sensing data of an environment; the data preprocessing module is connected with the multi-modal data acquisition module and is used for preprocessing the multi-modal sensing data; the feature extraction module is connected with the data preprocessing module and used for extracting features from the preprocessed multi-modal sensing data; the dynamic space-time diagram construction module is connected with the feature extraction module and is used for constructing a dynamic space-time diagram based on the features; the multi-modal data fusion module is connected with the dynamic space-time diagram construction module and is used for fusing the multi-modal data in the dynamic space-time diagram; through adaptive time alignment and a cross-modal space attention mechanism, the system can effectively process space-time inconsistency of data of different sensors, and centimeter-level positioning precision is realized.

The invention provides a landslide displacement space-time prediction method, device and system and a storage medium, and relates to the technical field of engineering geology, and the landslide displacement space-time prediction method comprises the steps: obtaining the historical landslide multi-field time sequence monitoring data of each monitoring point of a to-be-measured region, constructing a monitoring point bitmap based on the historical space related data, constructing a time sequence characteristic matrix according to the historical displacement time sequence, the historical landslide external disturbance factor time sequence and the monitoring point bitmap; inputting the time sequence characteristic matrix and the monitoring point bitmap into a trained landslide displacement space-time prediction model to obtain a landslide displacement space-time prediction result; the landslide displacement space-time prediction model is constructed based on an improved GCN network and an improved BiLSTM network; according to the method, the space-time prediction precision of landslide displacement can be remarkably improved by combining the space correlation data and the external disturbance factors and utilizing the capability of the deep learning model.

The invention discloses an establishment method of a rock multi-field coupling model and the model thereof, and belongs to the technical field of engineering geological survey. Through comprehensive application of multi-scale modeling, a nondestructive testing technology, a machine learning algorithm and intelligent optimization, the problems that existing multi-physical field coupling analysis is not deep enough, the precision of an observation method is poor, and a device is prone to being damaged are solved, the advantages of various observations are combined, the building and analysis precision of a rock multi-field coupling model is improved, and the rock multi-field coupling analysis precision is improved. The method has the advantages that the method is simple and convenient to operate, the intelligence and automation level of the model is enhanced, microstructure data and macroscopic response are combined, deeper understanding of rock behaviors is provided, meanwhile, more accurate and efficient decision support can be provided for rock engineering design and construction, and the method is worthy of popularization and application. And a novel analysis and simulation method is provided for the field of rock mechanics research. Through the comprehensive scheme, behaviors of rocks under complex geological conditions can be better predicted and managed, so that the safety and economical efficiency of engineering are improved.

A terahertz electromagnetic metamaterial multi-frequency wave-absorbing characteristic analysis design method based on deep learning and structural parameters comprises the steps that (1) a structural prototype and a parameterized model of a terahertz electromagnetic metamaterial are established, and an electromagnetic metamaterial data set containing terahertz multi-frequency wave-absorbing characteristic information is constructed through iterative simulation calculation; 2) segmenting the data set into a 80% training data set and a 20% test data set; 3) establishing bidirectional spectrum networks SN-1 and SN-2, sending the training data set into the SN-1 and SN-2 for training, and verifying the performance through the test data set; 4) inputting the training data set into SN-1 and SN-2 network training, and testing the training effect by using the test data set; 5) utilizing the multivariate Gaussian distribution curve to randomly generate double-peak, three-peak and four-peak spectrum curve data as a multi-peak wave absorbing curve needing to be designed; and 6) inputting spectral curve data into an SN-2 network for retrieval, reconstructing a terahertz electromagnetic metamaterial multi-frequency wave-absorbing structure in CST and performing simulation calculation, and designing an electromagnetic metamaterial structure with a multi-frequency wave-absorbing characteristic.

The invention belongs to the technical field of water quality prediction, and particularly relates to a river basin water quality prediction and early warning method and device based on a multi-model processing technology and a medium, and the method comprises the steps: firstly obtaining and preprocessing historical water environment monitoring data and real-time water environment monitoring data of a river basin; then constructing a water quality prediction deep learning model, and outputting a river basin water quality prediction result after training verification; constructing a water quality index prediction mechanism model, and after training verification, outputting water quality index prediction results of the river basin according to water quality prediction results; and finally, performing threshold judgment based on the river basin water quality prediction result and the dynamic change result of each water quality index of the river basin, and calling an expert emergency early warning plan according to the judgment result. According to the invention, the problem of poor prediction effect in the prior art can be solved.

The invention relates to the technical field of numerical simulation, in particular to a grid model generation method for the fluid displacement process of well cementation engineering. The method comprises the following steps: acquiring geological data of a well cementation project area, and constructing an initial geological model; well cementation engineering design data are obtained, well cementation engineering topological structure modeling is carried out, and a well cementation engineering topological structure model is obtained; according to the initial geologic model and the well cementation engineering topological structure model, fluid dynamic simulation is carried out, and a fluid characteristic matrix is obtained; performing fluid displacement process simulation according to the fluid characteristic matrix to obtain fluid displacement process data; performing fluid displacement process optimization on the well cementation engineering design data based on the fluid displacement process data to obtain optimized fluid displacement process data, and performing resource allocation redistribution on the well cementation engineering design data according to the optimized fluid displacement process data to obtain optimized well cementation engineering design data. According to the invention, the accuracy of fluid displacement simulation can be improved.

The invention discloses a physical knowledge-guided icing depth migration prediction method in a small sample scene, and the method comprises the steps: obtaining small sample icing data of a first region as a target domain, and obtaining large sample icing data of a second region as a source domain; constructing a physical information neural network and physical constraint loss thereof, and training the physical information neural network by using the source domain to predict the output icing thickness; a physical information neural network and self-migration learning are fused, knowledge of a source domain is effectively migrated to a target domain under guidance of physical knowledge, icing depth migration is carried out through iteration, an icing depth migration prediction model guided by the physical knowledge in a small sample scene is established, and the icing thickness of a first area is predicted. Physical knowledge is embedded into a transfer learning framework, a physical information neural network and transfer learning are fused, knowledge distillation loss based on cross entropy is used, and the physical knowledge in a teacher model is transferred into a student model, so that the student model adapting to a target domain is obtained.

The invention belongs to the technical field of medicine, and discloses a method and system for predicting anastomotic fistula occurring after colorectal cancer operation, and the method comprises the steps: obtaining clinical pathology data and laboratory analysis data of a patient; filling missing data in the clinical pathology data and the laboratory analysis data; screening key indexes by using a single-factor regression analysis technology; performing multi-factor regression analysis on the key indexes, and screening to obtain independent prediction factors for predicting whether the colorectal cancer patient has the anastomotic fistula or not after operation; based on the independent prediction factors obtained through screening, an anastomotic fistula prediction model is constructed and trained, and the trained anastomotic fistula prediction model is used for predicting the postoperative anastomotic fistula occurrence probability of the colorectal cancer patient. The method can more accurately predict the probability of occurrence of anastomotic fistula of a colorectal cancer patient after an operation, has better accuracy and reliability, can facilitate preoperative prediction of the probability of occurrence of anastomotic fistula for clinical doctors, and assists in guiding treatment decision.

The invention relates to the technical field of water conservancy projects, in particular to a water conservancy project digital management method based on a BIM model. The method comprises the following steps: acquiring real-time monitoring data; carrying out high-precision topographic data acquisition on a water conservancy project area, and carrying out data preprocessing to obtain preprocessed data; according to the preprocessed data, hydrological information integration modeling is carried out by utilizing a BIM platform, and a hydrological information integration model is obtained; performing multi-dimensional terrain dynamic updating on the hydrological information integration model by using the real-time monitoring data to obtain digital terrain data; performing flood simulation calculation based on the digital topographic data to obtain a flood simulation calculation result; and performing flood scene simulation according to the flood simulation calculation result to obtain flood simulation data. Through real-time monitoring, accurate prediction and timely early warning of the risk of the water conservancy project, a more effective technical means is provided for flood control and disaster reduction and safety management of the water conservancy project.

The invention is suitable for the field of pressure storage power station construction monitoring, and particularly relates to a BIM-based pressure storage power station underground chamber construction monitoring method and system, and the method comprises the steps: obtaining monitoring data used for representing the surrounding rock state of an underground chamber in a construction model; extracting feature distribution of the monitoring data in a time dimension by using a long-short term memory neural network; extracting feature distribution of the monitoring data in a spatial dimension based on a graph attention network; fusing the feature distribution of the monitoring data in the time dimension and the space dimension by adopting a cross attention fusion network structure, and constructing fusion features of the monitoring data; and taking the fusion features as input of a pre-constructed prediction model, and outputting a surrounding rock state prediction result. According to the invention, the BIM model and the monitoring data are combined, so that the construction state of the underground chamber can be comprehensively understood; and on the basis of feature extraction of monitoring data, monitoring and early warning of the surrounding rock state are realized, and the technical problem of poor accuracy of a monitoring result of a traditional monitoring method is solved.

The invention discloses an intelligent operation and maintenance management method and system based on an artificial intelligence data center, and relates to the technical field of data processing, and the operation and maintenance management system comprises a data collection module, a data classification storage module, a data prediction module, a data processing module and a data evaluation module. The data acquisition module is used for acquiring different hardware equipment data and software data in the operation and maintenance management system in real time. The data classification storage module is used for classifying and storing the data in the operation and maintenance management system, and then the importance of different data is determined by utilizing grades; and then the data prediction module analyzes the classified data, so that the operation and maintenance management system can predict different classification data conveniently, after fault data occurs in the operation and maintenance management system, judgment is performed between the fault data and the classification, and the position of the fault data in the operation and maintenance management system is effectively and quickly positioned.

The invention discloses a self-adaptive Monte Carlo macro particle simulation method and system based on parallel computing, and relates to the technical field of particle simulation, and the method comprises the steps: initializing parameters, if the current step number is not the maximum step number, storing node data, balancing particle loading, and calculating charge density and battery density; using macro particles to represent a group of real particles, updating the positions and speeds of all the particles in parallel, and calculating new charge distribution; setting adaptive grid configuration, and dynamically adjusting grid density according to electric field intensity and local density of charge distribution; and based on the updated particle position and speed, evaluating and processing a collision event, storing collision data, re-calculating surface charge density data, adding 1 to the current step number, and circularly executing to finish simulation when the maximum step number is reached. According to the method, the accuracy of a simulation result is improved through the improved collision model, electrical discharge behaviors under different conditions are better predicted, and more reliable data support is provided for design and safety evaluation of high-voltage equipment.

The invention discloses a flue gas outlet NOx concentration prediction method based on LSTM model real-time calibration. The flue gas outlet NOx concentration prediction method comprises the following steps: step 1, determining depth related factors; step 2, data acquisition; step 3, data preprocessing; step 4, data division; step 5, constructing an LSTM (Long Short Term Memory) model; 6, performing model training and precision evaluation; step 7, predicting the NOx concentration of the flue gas exhaust port; and 8, calibrating the model in real time. According to the method, accurate prediction of the NOx concentration of the flue gas outlet can be realized; the problem of large lag existing in outlet NOx concentration measurement can be solved, accurate ammonia spraying is realized, and the consumption of a denitration reducing agent is reduced; the method has good generalization ability and strong applicability, and is convenient to popularize in practical application; compared with a traditional statistical model, the method has the advantages that the NOx emission concentration can be predicted more accurately, and a more reliable decision basis is provided for environmental protection and pollution abatement.

The invention relates to the field of pedestrian trajectory prediction, in particular to a pedestrian trajectory prediction method, device and equipment based on a progressive refinement strategy. The prediction method comprises the following steps: predicting a current historical trajectory by using a trained prediction model, and outputting a future prediction trajectory; wherein the trained prediction model is obtained by training a built model through a training data set, the training data set comprises multiple groups of trajectory data, and each group of trajectory data comprises a historical trajectory X and a future real trajectory Y; the model comprises a motion mode clustering block, a self-adaptive motion mode refining block, an embedded splicing block, a Transform encoder, a Transform decoder, a track refining block and a classifier. According to the method, the motion mode and the prediction track are gradually refined through the two refining blocks, the adaptability of the model to different scenes is enhanced, and the prediction of complex motion is improved.

The invention discloses a shield tunnel settlement prediction method based on artificial intelligence. The method comprises the steps of data acquisition, data optimization, construction of a shield tunnel settlement prediction model, model optimization and settlement prediction. The invention relates to the technical field of tunnel construction monitoring, in particular to a shield tunnel settlement prediction method based on artificial intelligence. According to the scheme, data cleaning is carried out by defining a dimension refining distance function and calculating gravitation between data, potential relations between the data are deeply mined, noise interference is reduced, and the accuracy and reliability of the data are improved; by designing a mapping function, a circulation control layer, a depth correlation layer and a loss function, the recognition and prediction precision of the model on a complex settlement mode is improved; model parameters are dynamically searched by designing an intervention factor, defining an adaptive step length, performing bidirectional iterative search and gradually adjusting search, so that the search efficiency is improved, time consumption and a loss value of model training are balanced, and the model is more efficient and stable.

The invention relates to a method for detecting a B cell subset in a lymphatic tissue, a kit comprising a quantum dot-antibody CD19 conjugate, a quantum dot-antibody CD138 conjugate and a quantum dot-antibody CD1d conjugate is used for dyeing a lymphatic tissue section, and different antibodies are marked by quantum dots with different colors, so that the B cell subset in the lymphatic tissue section can be detected, and the B cell subset in the lymphatic tissue section can be detected. Content and positions of plasma cells and regulating B cells are confirmed, so that biological behaviors of the plasma cells and regulating B cells can be determined conveniently. According to the method, the existing multi-color immunofluorescence technology is improved, a secondary antibody dyeing link is omitted by using a quantum dot carrying antibody method, a direct method is adopted for marking, the advantages of quantum dots are exerted, the experiment time is saved, and the efficiency is improved.

The invention relates to an intelligent fund routing system based on an AI large model and an optimization method thereof, and belongs to the field of finance, and the system comprises a data processing module which is used for automatically collecting financial market data, and carrying out the preprocessing and feature extraction; the decision forming module is used for analyzing the processed data and performing fund routing decision based on an AI model architecture; the fund routing optimization module is used for optimizing a fund routing path according to market and user behavior prediction results; the risk assessment module is used for monitoring transactions and fund flow directions in real time and assessing potential risks; and the risk management integration module is used for carrying out early warning and decision support according to the risk assessment result. The method has the advantages that the efficiency is improved; the safety is enhanced; the precision is improved, and the deep learning model can better predict the market trend and the user behavior.

The invention relates to the technical field of data analysis, in particular to an innovation achievement evaluation system based on big data, and the system comprises a data capturing module which screens an original data set of an enterprise scientific and technological innovation project, extracts time series data, and distinguishes the data into cost data, an innovation project milestone date and a quality index; the cost change rate, the project progress deviation and the quality change index of the multiple innovation projects are calculated, the average value and the standard deviation of multiple indexes are counted, and preliminary project deviation analysis is established; according to the method, time sequence data of science and technology projects are accurately analyzed, cost, progress and quality indexes are monitored and analyzed, time sequence decomposition and independent analysis of seasonal and trend deviation are carried out, the recognition precision of periodic changes and long-term development trends of the projects is improved, key data changes can be accurately captured, and the recognition accuracy is improved. Accurate positioning and prediction of key change points in scientific and technological progress are promoted, and evaluation and optimization effects on research and development achievements are greatly enhanced.

The invention discloses a bullfrog biological behavior recognition optimization system based on deep learning, and relates to the technical field of biological recognition. Performing frame decomposition processing on the image data of the bullfrogs to obtain image data frames, extracting first-class feature information and second-class feature information in the image data frames, training an image recognition model to construct moving tracks of different bullfrogs, obtaining theoretical active coefficients and correlation coefficients of the bullfrogs according to the moving tracks, and obtaining the theoretical active coefficients and the correlation coefficients of the bullfrogs according to the theoretical active coefficients and the correlation coefficients; optimizing the theoretical activity coefficient to obtain an actual activity coefficient, inputting the actual activity coefficient into a preset behavior prediction model, obtaining a predicted movement track of the bullfrog, and obtaining predicted bullfrog density of different activity areas according to the predicted movement track; the bullfrog activeness monitoring accuracy can be improved, and the bullfrog density prediction accuracy can be improved.

The invention relates to the technical field of sewage treatment, in particular to an intelligent operation management method and system for a water plant ultrafiltration membrane pool and a storage medium. The method comprises the steps that membrane pool operation parameters are obtained, multiple regression analysis is conducted on the membrane pool operation parameters, and transmembrane pressure differences corresponding to the membrane pool operation parameters are determined; the operating parameters of the membrane tank comprise inlet water quality, outlet water quality, water temperature and filtering pressure; when the transmembrane pressure difference is higher than the preset standard pressure difference, whether the current moment is in a preset emptying period or not is judged; if yes, carrying out cleaning operation based on the membrane pool operation parameters; and if not, obtaining a transmembrane pressure difference increase value in the first preset time period, and determining a coping intervention type based on the transmembrane pressure difference increase value, the coping intervention type including preventive intervention and immediate intervention. According to the invention, the balance between the cleaning and maintenance of the ultrafiltration membrane is conveniently improved, so that the continuous, efficient and stable operation of the water treatment process is ensured.

The present invention relates to the field of solid-state drive fault prediction, and in particular to a solid-state drive fault prediction method and system based on artificial intelligence. The method comprises the following steps: fine-grained division of the solid-state drive into storage units, and serialization of the storage units to construct a storage unit sequence for each partition; obtaining a solid-state drive operation log; performing multi-time point state evolution on the storage unit sequence for each partition based on the solid-state drive operation log, and performing deep dynamic evolution learning to generate a state evolution law for each storage unit; performing periodic access load peak calculation on the solid-state drive operation log, thereby generating a simulated access data packet for each storage unit; performing dynamic pressure access simulation based on the state evolution law of each storage unit and the simulated access data packet of each storage unit to obtain response state evaluation data for each storage unit. The present invention realizes efficient and accurate solid-state drive fault prediction.

The invention discloses a drug-target affinity prediction method based on a multi-scale convolutional neural network, and the method comprises the steps: obtaining a sequence of a target protein and an SMILES character string of a drug small molecule, obtaining an SMILES word vector matrix and a protein word vector matrix based on word vector coding, constructing the multi-scale convolutional neural network, carrying out the learning of the SMILES word vector matrix, and carrying out the prediction of the drug-target affinity. The method comprises the following steps: constructing a protein word vector matrix, acquiring chemical and structural characteristics of drug small molecules, constructing a one-dimensional convolutional neural network, learning the protein word vector matrix, acquiring structural characteristics of target protein, and fusing the chemical and structural characteristics of the drug small molecules and the structural characteristics of the target protein through splicing operation. And expressing the fused features as drug-target features, constructing an affinity prediction model, and learning the drug-target features to obtain an affinity prediction value between the drug and the protein. Compared with a graph neural network and a pre-training model, the drug-target affinity can be better predicted, and the prediction capability is better than that of a deep learning method.

The invention discloses a COD (Chemical Oxygen Demand) prediction method based on a two-channel convolution time sequence mechanism, which comprises the following steps of: acquiring COD original data for preprocessing to obtain COD input characteristics; constructing a prediction model, and importing data into a dual-channel convolution module; importing the first double-channel convolution output feature vector and the second double-channel convolution output feature vector into a double attention memory network to obtain a time sequence dependence feature; the time sequence dependence characteristics are imported into an output module, and predicted COD output is obtained; training the model, and in the training process, optimizing hyper-parameters of the model based on an improved pelican algorithm; and importing real-time COD original data into the trained model, and obtaining real-time COD output. According to the method, a high-efficiency effluent COD content prediction model of the sewage treatment plant is constructed by creatively fusing the two-channel convolution and the double-attention memory network.

The invention discloses an application method and system of an active light-emitting sign based on a road environment and traffic flow, and the method comprises the steps: collecting road environment data and traffic flow data in real time through employing a vehicle-mounted sensor and a roadside monitoring device, and the road environment data comprises the temperature and real-time illumination of an area where a road is located; transmitting the road environment data and the traffic flow data to a traffic control center; the traffic control center calculates the brightness, the color and the flicker frequency required by the LED lamp bead of the active light-emitting sign device in real time; transmitting the calculated brightness, color and flicker frequency to the active luminous sign equipment through wireless communication; the microcontroller of the active light-emitting sign device adjusts the LED lamp beads in real time according to the received brightness, color and flicker frequency. The method and the system can intelligently adapt to different weather, time and road conditions, are energy-saving and environment-friendly, are high in safety, are beneficial to traffic management optimization, and are high in expandability.

The invention provides a water environment index change trend prediction and model establishment method and device, and a medium, and the model establishment method comprises the following steps: obtaining historical index data of a target water environment and driving influence factor data of the historical index data, and enabling the historical index data and the driving influence factor data to be used for establishing a training data set; hyper-parameters of the water environment index change trend prediction model are determined by adopting an automatic optimization algorithm, and the key hyper-parameters comprise a learning rate, the size of a convolution kernel and the number of neurons; the training data set is used for training a water environment index change trend prediction model, and a basic network model of the water environment index change trend prediction model is a combination of a convolutional neural network, a long and short term memory neural network and a self-attention mechanism. The method is based on the CNN-LSTM-Attention neural network model and adopts the automatic optimization algorithm to adjust the hyper-parameters, so that the parameter setting optimality in the model training process is ensured, higher adaptability and stability are obtained, and a more reliable water environment index change trend prediction result can be provided.

The invention discloses a potential risk hidden danger identification method and device based on multi-dimensional information and event causal sorting, equipment, a medium and a product, and relates to the field of artificial intelligence. The method comprises the following steps: acquiring multi-modal data from a subscribed public opinion monitoring object, performing feature extraction, and extracting multi-modal information features; performing data structured analysis based on the multi-modal information features, and extracting key attribute structured data of each event; performing clustering analysis on all the events, and dividing all the events into a series of large event sets; performing information deduplication and integration analysis on each small event in the large event set to obtain key attribute structured data of each large event set, performing causal chain sorting on an occurrence sequence and a causal relationship of a series of large event sets to obtain an event causal chain, and performing potential risk hidden danger identification on the event to obtain a hidden danger identification result. The comprehensiveness and timeliness of public opinion monitoring and the accuracy of public opinion analysis can be improved.

The invention discloses an inertia and damping adaptive VSG control method based on neural network optimization. The method comprises the steps that S1, a VSG rotor motion equation is obtained through equivalent conversion; s2, introducing a speed regulator into the active power-frequency control loop to obtain a VSG reactive power-voltage control equation; s3, performing circuit model equivalence on the VSG grid-connected system to obtain an output active power equation, and obtaining a relation between inertia and damping; s4, analyzing power angle characteristics and angular frequency oscillation under a disturbance condition, and summarizing an inertia self-adaptive mechanism; and S5, establishing an artificial neural network, analyzing and predicting frequency and voltage changes of the power grid, and adjusting weight values to realize self-adaptive adjustment of inertia and damping. According to the inertia and damping adaptive VSG control method based on neural network optimization provided by the invention, the inertia and damping of the system can be cooperatively adjusted, the frequency is stabilized, the overshoot is reduced, and the method has faster response time in the aspect of recovering to a stable state.

The invention relates to the technical field of electric digital processing, in particular to a multi-modal knowledge graph completion method based on composite operation. The method comprises the following steps: acquiring a multi-modal knowledge graph, generating text description through a picture, and converting the multi-modal knowledge graph into a text knowledge graph comprising a plain text tetrad; learning a first embedded representation of the plain text tetrad based on a knowledge graph embedding model; learning a second embedded representation of the multi-modal tetrad based on a feedforward neural network model; inputting the current head entity, the current time and the current relationship into the knowledge graph embedding model and the feedforward neural network model to obtain a predicted tail entity; and complementing the multi-modal knowledge graph according to the predicted tail entity. According to the invention, high-precision feature expression is carried out in the multi-modal knowledge graph, so that complementation of the multi-modal knowledge graph is completed.

The invention relates to the field of robot operation and deep learning model algorithm design, in particular to a robot operation method and system based on a waypoint perception world model.The robot operation method comprises the steps that firstly, operation information of a robot is collected, and the operation information comprises an observation image, current state information of the robot and an instruction given by a user; the operation information is input into a preset waypoint sensing world model, the model generates waypoint information of the robot based on input, the waypoint information and state information of the robot are used for calculating motion information of the robot through a preset motion prediction module, the calculated motion information of the robot is combined with the waypoint information, and the motion information of the robot is obtained. According to the method, the waypoint sensing world model captures key information, and good generalization ability is kept in a changing scene.

The invention provides a runoff real-time rolling forecasting method, which comprises the following steps of: performing runoff forecasting by constructing a distance weighted self-regularization neural network model, and inputting training set data into the distance weighted self-regularization neural network model after data preprocessing; according to the distance weighted self-regularization neural network model, multiple groups of LSTM and CNN neural networks are used for learning training set data from three directions of points far away from a mean value, points close to the mean value and index data, then an FC layer is used for data fusion, loss is calculated through multiple distance weighted loss functions, and finally a forecast result is obtained. In the forecasting process, a sequence of a time step length is forecasted backwards based on the current time, and the future runoff process can be forecasted day by day in a rolling manner by setting the forecasted time step length and continuously updating input data, so that real-time forecasting is realized; and collecting a first time period prediction result of each time sequence, and taking a merged result as a runoff prediction process of the whole prediction period, thereby realizing a real-time prediction effect.