373 results about "Recurrent neural network model" patented technology

The invention discloses an underground construction decision-making method based on three-dimensional geological modeling and risk hot area identification, and relates to the field of fusion of artificial intelligence and geological engineering. The method comprises the following steps: firstly, acquiring drilling data, geological radar images and seismic reflecting layer information, constructing a three-dimensional geological voxel model with spatial topology constraints, and accurately describing a geological unit structure by adopting an irregular grid mode; and then, extracting a time sequence characteristic index under construction disturbance, forming a continuous time sequence characteristic vector, inputting the continuous time sequence characteristic vector into a convolutional recurrent neural network model with a space attention aggregation mechanism and a deep memory unit, and predicting a risk heat value of each space position. And on the basis, through heat gradient clustering and neighborhood consistency analysis, a dynamic high-risk hot area is identified, and a risk hot area map is constructed. And finally, in combination with the construction stage, the equipment plan and the sensor feedback information, constructing a multi-target auxiliary decision function, and generating a construction decision result including operation path reconstruction, rhythm adjustment and power limit and control suggestions.

The invention relates to the technical field of battery thermal management, in particular to a battery thermal management control method and system, and the method comprises the steps: obtaining the basic operation parameter data of an automobile battery in real time, and carrying out the multi-dimensional data fusion, and obtaining a multi-dimensional data set; based on the thermodynamic model and the multi-dimensional data set, the temperature distribution condition in the battery is calculated in real time, and real-time temperature distribution information is obtained; based on the historical data and the operation state, the future temperature change trend of the battery is predicted through a recurrent neural network model; dynamically adjusting a control strategy of the battery thermal management system according to the real-time temperature distribution information and the prediction result; and analyzing the temperature difference condition of the battery monomers, detecting abnormal temperature rise, and performing temperature difference balance adjustment and triggering a multi-stage protection mechanism when the abnormal temperature rise exceeds a threshold value. According to the invention, the problems of inaccurate temperature prediction and response lag in the prior art are effectively solved, the control strategy is dynamically adjusted through real-time data fusion and recurrent neural network prediction, temperature difference balance and abnormal temperature rise detection are realized, the safety of the battery is improved, and the service life of the battery is prolonged.

The invention discloses a nuclear radiation index early warning system based on multi-algorithm fusion, and the system comprises an environment data collection module which is used for collecting gamma-ray dose rate data, wind speed data, wind direction data and geographic position information, and constructing a nuclear radiation environment data set; the dose rate prediction module is used for predicting a future change trend of gamma ray dose rate data based on a sparse recurrent neural network model; the anomaly detection module is used for identifying an abnormal mutation point and positioning the position of a high-risk sensor; the diffusion region modeling module is used for calculating a nuclear radiation diffusion influence region; the grid risk evolution module is used for simulating time sequence evolution of a regular grid risk state based on the multi-state cellular automaton; and the fusion judgment module is used for fusing output results and generating a final early warning grade and a spatial risk distribution result. According to the invention, accurate prediction of the nuclear radiation risk and dynamic space early warning are realized, and the early warning accuracy and response efficiency are significantly improved.

The invention relates to the technical field of agricultural product circulation management and traceablility, in particular to an agricultural product circulation traceablility method and system.The method comprises the steps that environment states such as temperature, humidity and illumination are recorded in real time through an intelligent material label and dynamically bound to a digital twinborn body; an improved isolated forest algorithm is combined to detect environment abnormity, and an abnormity detection result is generated; performing trend prediction on the complemented time sequence data by using a recurrent neural network model and a multivariable interpolation algorithm to generate future state data; and constructing a knowledge graph containing historical, current and future states based on the digital twinborn body and the prediction data, and generating a traceability path and a risk prompt. According to the invention, real-time monitoring, accurate prediction and intelligent traceability of the agricultural product circulation process are realized, the risk early warning capability and data transparency are improved, and reliable decision support is provided for consumers, producers and supervision mechanisms.

The invention discloses a health risk assessment method and early warning system based on multi-source data analysis. The health risk assessment method comprises the following steps: S1, collecting and preprocessing multi-source health data through medical detection equipment; s2, extracting key health indexes, sequence features and statistical features based on the health data set; s3, adopting a recurrent neural network model to construct a health risk assessment model; s4, optimizing structural parameters of the health risk assessment model by adopting an improved dragonfly algorithm; s5, performing performance evaluation on the health risk evaluation model by using the optimal parameter set; s6, deploying the final health risk assessment model in a health risk assessment and early warning system; and S7, when the health risk assessment result exceeds a preset risk threshold, automatically generating early warning information. According to the method, the recurrent neural network model, the improved dragonfly algorithm and the multi-source health data fusion optimization technology are combined, and health risk assessment and early warning based on multi-source data analysis are realized.

The invention relates to the technical field of isolator service life prediction, and discloses an isolator remaining service life prediction method based on deep learning. The method comprises the following steps: acquiring a vibration signal and a temperature signal when the isolator operates, and constructing a multi-dimensional sensor sequence; dividing a plurality of sub-sequence units containing a fixed time window; extracting time domain and frequency domain features of each sub-sequence unit, and generating a fusion feature vector; calculating statistical distribution characteristics and variation trend characteristics of key indexes in the vector, and outputting a health state index; through a bidirectional recurrent neural network model containing an attention mechanism, taking the health state index and the historical degradation data as input, processing a time sequence dependency relationship through multi-layer residual error connection, and generating a residual service life prediction value; and updating the multi-dimensional sensor sequence according to the real-time sensor signal, and dynamically correcting the predicted value. According to the method, multi-dimensional data are integrated, degradation characteristics are deeply mined, and the accuracy and adaptability of prediction of the remaining service life of the isolator can be improved.

Provided in the present invention are a power-distribution-network self-healing method and system taking photovoltaic output into consideration. The method comprises: acquiring historical operation data of a photovoltaic power station and irradiance observation data from a meteorological station; on the basis of the historical operation data of the photovoltaic power station and the irradiance observation data from the meteorological station, predicting the generated power of the photovoltaic power station by using a convolutional long-short-term memory recurrent neural network model that takes sparrow search into consideration; on the basis of the generated power of the photovoltaic power station, a segment-switch state of a power distribution network and a network topology of the power distribution network, constructing an objective function and a constraint condition for a power-distribution-network self-healing model, and obtaining the power-distribution-network self-healing model; solving the power-distribution-network self-healing model by using a propagation search algorithm, so as to obtain an optimal recovery strategy; and executing the optimal recovery strategy by means of segmented switches and node loads. The present invention can realize self-healing of a power distribution network while ensuring the minimum power generation cost of a distributed power source, the minimum network loss and the minimum node voltage deviation.

The invention discloses an unmanned aerial vehicle intelligent scheduling and management method based on task planning, and relates to the technical field of dynamic path planning, and the method comprises the steps: collecting flight data, and carrying out the preprocessing; based on the preprocessed data, constructing an intelligent patrol dynamic convolutional network model to analyze the health state of the unmanned aerial vehicle, and generating a health state result; performing dynamic path planning through a recurrent neural network model by using the generated health state result and combining environment perception data; making a task and executing a plan according to dynamic path planning and an intelligent scheduling algorithm; according to the method, hierarchical feature extraction and fusion are performed on the preprocessed flight data through the intelligent patrol dynamic convolutional network model, the health state feature vector is generated, and accurate health state evaluation and anomaly detection are realized in combination with the health state value and the One-Class SVM.

The invention discloses a classroom environment intelligent supervision method and system based on data analysis, and the method comprises the steps: obtaining multi-source heterogeneous data in a classroom environment, eliminating the sensor noise of the multi-source heterogeneous data through employing an improved Kalman filtering algorithm, carrying out the time-space alignment of the data after noise elimination, and obtaining the aligned multi-source heterogeneous data; establishing a CRNN three-layer convolutional recurrent neural network model, and inputting the aligned multi-source heterogeneous data into the CRNN three-layer convolutional recurrent neural network model for recognition to obtain a classroom teaching environment state; and carrying out classroom risk early warning and self-adaptive classroom environment regulation and control according to the classroom teaching environment state. The classroom emergency risk can be actively prevented, automatic adjustment can be achieved through an intelligent air conditioner and an illumination system, a safe and comfortable teaching environment is created, and classroom efficiency improvement and teacher and student health guarantee are facilitated.

The invention belongs to the technical field of pollution source detection, and discloses a pollution source detection method and system for water environment treatment. Comprising the steps that acoustic disturbance data are collected through an underwater acoustic sensor, real-time edge preprocessing is carried out, water environment data of different depths are synchronously collected, water environment gradient data are calculated, and then data synchronization is achieved through a unified timestamp. Fourier transform is adopted to extract frequency domain characteristics of the sound disturbance signals, the main frequency, the energy difference and the frequency spectrum skewness are calculated, and disturbance sound wave indexes are generated according to the main frequency, the energy difference and the frequency spectrum skewness. In combination with disturbance sound wave indexes and water body environment gradient data, a sliding window is adopted to construct a time sequence sample set, a recurrent neural network model for predicting future pollution release risks is trained, differentiated pollution release threshold values are set according to different water body types, and when model output exceeds the threshold values, sediment pollution early warning instructions of corresponding levels are automatically generated. And efficient and graded pollution event intelligent identification and response are realized.

According to the hydrogen clock error forecasting method, system and device based on the improved recurrent neural network and the storage medium, the memory characteristic of the recurrent neural network is used for representing the time relation, so that hydrogen clock error forecasting is carried out, and the accuracy of hydrogen clock error forecasting is improved. The method comprises the following steps: acquiring original clock error data of a hydrogen clock, and preprocessing the original clock error data to obtain preprocessed clock error data; constructing a hybrid recurrent neural network model, and initializing the hybrid recurrent neural network model; dividing the preprocessed clock error data into a training set, a verification set and a test set; based on the training set, the verification set and the test set, using an RAdam optimizer and MAPE as loss functions, and optimizing the hybrid recurrent neural network model until a preset precision is reached; and preprocessing the clock error data of the to-be-predicted time period, and inputting the preprocessed clock error data of the to-be-predicted time period into the optimized hybrid recurrent neural network model to obtain a predicted clock error value.

The invention belongs to the field of distributed optical fiber sensing, and provides a Brillouin temperature extraction method and system based on a self-attention recurrent neural network, and the method comprises the steps: collecting real Brillouin gain spectrum data through a Brillouin optical time domain analysis system which is built in advance, and carrying out the preprocessing of the real Brillouin gain spectrum data; based on the preprocessed Brillouin gain spectrum data, utilizing the trained self-attention recurrent neural network model to obtain a corresponding Brillouin temperature; a meta-learning algorithm is introduced in the process of a self-attention recurrent neural network model, so that the network model can share initialization parameters among a plurality of tasks, the parameters are optimized through meta-training, the network model can quickly adapt to a new task, training from zero is not needed in the face of multi-task learning or the new task, and the training efficiency is improved. The model can be quickly adjusted through a small amount of gradient updating, the training time is shortened, the performance is improved, and the problems of long time, poor real-time performance, poor adaptability and the like of curve fitting are solved.

The invention discloses an investment risk early warning method and system for the whole life cycle of a power grid, and relates to the technical field of power system risk management, and the method comprises the steps: collecting historical operation data and real-time data, carrying out the preprocessing, constructing a project feature vector matrix P, inputting a recurrent neural network model for training, and obtaining a risk recognition reference model M; inputting the real-time data into the risk identification reference model M, and extracting risk feature values R of the power grid investment project in different stages; performing clustering analysis by using a fuzzy c-means clustering algorithm to obtain a risk grade division threshold T; and inputting the risk characteristic value R and the risk grade division threshold T into a constructed risk evolution model E, calculating to obtain a risk evolution trend index alpha and a risk mutation probability beta, inputting the risk evolution trend index alpha and the risk mutation probability beta into a Bayesian early warning network for early warning, and feeding back an early warning result to the risk identification reference model M for updating. According to the invention, through data driving and model adaptive improvement, the intelligence of risk early warning is realized.

The invention discloses a photovoltaic power generation power prediction method based on Bi-LSTM, and relates to the technical field of photovoltaic power generation power prediction methods. The method comprises the following steps: constructing a Bi-LSTM recurrent neural network model; historical data of photovoltaic power generation power and related meteorological data are obtained to serve as model data and are aligned into a time sequence; preprocessing the model data and training a Bi-LSTM recurrent neural network model; and optimizing hyper-parameters of the Bi-LSTM recurrent neural network model by using a genetic algorithm to obtain an optimal Bi-LSTM recurrent neural network model to predict photovoltaic power generation power. According to the method, the Bi-LSTM model is utilized, and past and future information in the time sequence data can be captured at the same time, so that the accuracy of photovoltaic power generation power prediction is improved; meanwhile, effective features can be screened and utilized more effectively, the defects that a traditional prediction model is poor in feature screening capacity and prone to noise interference are overcome, and then the prediction precision of the photovoltaic power generation power is remarkably improved.

The invention relates to the technical field of ultrasonic detection and noise reduction, and provides an external transmission hose defect ultrasonic detection method based on reciprocity and a hybrid neural network, and the method comprises the steps: carrying out the filtering processing of an initial signal, and obtaining a first signal; decomposing the first signal into an approximation coefficient and a first detail coefficient; performing soft threshold processing on the first detail coefficient to obtain a second detail coefficient, and reconstructing the second detail coefficient and the first approximation coefficient to obtain a second signal; performing normalization processing on the second signal to obtain a third signal; inputting the third signal into a convolutional neural network model to obtain a spatial feature vector; inputting the spatial feature vector into a recurrent neural network model, and extracting a time feature vector; fusing the spatial feature vector and the time feature vector to obtain a noise reduction signal; the convolutional neural network model and the recurrent neural network model are trained and then integrated with an ultrasonic detection system, defect characteristic signals are output, and the online detection precision and detection efficiency of the output hose are improved.

The invention relates to the technical field of robot control, and discloses a mechanical arm trajectory planning control method and system based on a BAFARNN model. The method comprises the steps that a mechanical arm kinematics model is established, and a trajectory tracking problem is converted into a time-varying equation; designing a bounded adaptive function to activate a recurrent neural network model, defining an error function and constructing a dynamic equation; designing a piecewise adaptive coefficient function, and dynamically adjusting the gain according to an error norm and time; setting a Lissajous curve as an expected trajectory, and initializing a simulation environment; the joint speed is solved in real time through an ODE numerical method, and the mechanical arm is driven to move; actual motion data is collected and compared with an instruction, and closed-loop feedback control is triggered when the actual motion data exceed a threshold value. According to the method, rapid convergence is achieved through the piecewise adaptive coefficient function, the bounded activation function and the negative feedback mechanism are adopted to suppress noise, and high-precision and real-time trajectory tracking of the mechanical arm in the dynamic environment is achieved.

The invention discloses a photovoltaic energy storage control method and a photovoltaic energy storage control system, and relates to the field of G06F30 / 27 machine learning. According to the method, a photovoltaic energy storage system trains a preset convolutional neural network model and a recurrent neural network model to obtain an optimal working point calculation model and a model for predicting electricity consumption in one hour in the future; the system obtains the current maximum power tracking point parameter and the electricity consumption in one hour in the future according to different data input models by using data collected in real time; when the energy storage proportion value of the predicted total electric quantity and the electricity consumption in one hour in the future does not exceed a preset threshold value, controlling the photovoltaic panel to generate power according to the maximum power tracking point parameter; if yes, the power generation duration proportion is controlled according to the proportion. By implementing the technical scheme provided by the invention, the power generation behavior of the photovoltaic panel is flexibly adjusted according to real-time power generation and power consumption conditions, so that the waste of electric energy is avoided, stable power supply of the system under different power consumption requirements can be ensured, and the adaptability and flexibility of the whole photovoltaic energy storage system are improved.

The invention relates to the technical field of resource scheduling, in particular to a digital-intelligent non-accompanying service management scheduling system, which comprises a service baseline construction module for instantiating a service item into a standard plan event flow; the state portrait generation module is used for collecting service object data flow, establishing a dynamic time sequence signal by applying a recurrent neural network model, and quantifying the occurrence probability of a future abnormal event through multi-step prediction; when the triggering is abnormal, executing causal two-way collision positioning reasons by applying the knowledge graph; the task load calculation module is used for integrating the plan event flow into a plan task component, and generating a worker resource list in combination with the predictive risk score and the worker real-time load and space reachability analysis containing the fatigue index; the system further comprises a task distribution module, the optimal staff is determined and a closed-loop scheduling instruction is generated by minimizing a composite cost function containing predictive load cost, and digital intelligent unattended service management from prediction to scheduling is realized.

The invention provides a parallel computing method for supporting real-time data stream processing in parallel programming, which comprises the following steps of: dividing a real-time data stream into fragments through a parallel pipeline, and filtering specified fields by different threads to improve the efficiency; after filtering is completed, performing local aggregation on a result by using a local aggregation function of MapReduce to reduce the data volume; a distributed memory database based on Redis Cluster is established, and data are stored in a fragmented mode according to a time window; constructing a long-short-term memory recurrent neural network model, and performing parallel training through a parameter server architecture; if the long short-term memory recurrent neural network model needs incremental learning, a parameter interface is reserved for iterative updating, and repeated training is avoided; after training is completed, the model is deployed to a GPU server, and real-time data flow is predicted; through a task scheduler of a distributed memory computing framework, cooperation between threads and processes is automatically managed, and programming work is simplified.

The invention discloses an atmospheric pollutant concentration prediction method of a self-organizing recurrent neural network based on Bayesian optimization, and belongs to the technical field of intelligent environmental protection. At present, a method for predicting the concentration of atmospheric pollutants has some problems including complex models, low prediction precision, insufficient space-time correlation, poor real-time performance and the like. Therefore, the invention designs a self-organizing recurrent neural network model based on Bayesian optimization. According to the model, the correlation between various meteorological factors and atmospheric pollutants is deeply mined by using the advantages of a gated circulation unit neural network; an intelligent tuning mechanism based on a Bayesian optimization algorithm is designed to realize preliminary optimization of model parameters and structures, so that the efficiency of the model is improved; and then the contribution degree of each neuron of a hidden layer to output is calculated by adopting sensitivity analysis, so that the network structure and parameters are adaptively adjusted, and the robustness and accuracy of the model are further improved.

The invention provides a map updating method and device, electronic equipment, a computer readable storage medium and a computer program product. The method comprises the steps of performing prediction processing based on historical change data of a global map and current environment data through a first convolutional neural network model to obtain a first prediction result of map elements in the global map; performing first fusion processing on the global map and the part of the local map at the plurality of collection times to obtain a fusion result of the plurality of collection times, and performing prediction processing based on the fusion result of the plurality of collection times through a recurrent neural network model to obtain a second prediction result of map elements in the global map; wherein the local map is determined according to the vehicle perception data of the vehicle driving on the global map; and performing map updating processing on the global map according to the first prediction result and the second prediction result of the map elements in the global map. According to the method, the timeliness and accuracy of map updating can be remarkably improved.

The invention discloses a lithium battery health state prediction method and system based on a neural network, and the method comprises the steps: obtaining multi-source heterogeneous data of a lithium battery, and decomposing the multi-source heterogeneous data; key features are obtained according to the multi-source heterogeneous data; constructing a training set, wherein the training set comprises multi-source heterogeneous data, decomposed data and key features; constructing a multi-scale recurrent neural network model, and training the multi-scale recurrent neural network model through the training set to obtain an optimized model; and acquiring operation data of the lithium battery, and predicting the operation data through the optimized model to obtain lithium battery health state sequence data in future time.

The invention discloses a distributed network node heartbeat message transmission optimization method and device. The method comprises the following steps: acquiring network link data among distributed network nodes; extracting a plurality of influence factors influencing the link communication of the distributed network nodes from the network link data, inputting the plurality of influence factors into the trained gated recurrent neural network model, and predicting the link communication state of the distributed network nodes and the adjustment strategy of the heartbeat message sending frequency through the gated recurrent neural network model; and according to a prediction result of the gated recurrent neural network model, the sending frequency of node heartbeat messages is adjusted, and a message transmission link of the distributed network nodes is optimized by using an ant colony algorithm. According to the method, the heartbeat detection frequency can be dynamically optimized under the environment of weak communication and limited bandwidth, the heartbeat detection efficiency is remarkably improved, and the network load is reduced.

The invention relates to the technical field of non-ferrous metal metallurgy, and discloses a multi-gun top-blown furnace high-temperature infrared monitoring spray gun position intelligent regulation and control method and system, and the method comprises the steps: carrying out the dimension reduction processing of key data in a blowing process through principal component analysis; processing the output preprocessed key data by using a recurrent neural network model; the recurrent neural network model captures a long-term dependency relationship of temperature change, predicts dynamic parameter change in the smelting process, and outputs an adjustment demand of the height of the spray gun; adjusting the height of the spray gun in real time by adopting a closed-loop control system based on the prediction result of the adjustment demand; and the predicted adjustment requirement of the spray gun height is converted into a control signal, and the spray gun height of the multi-gun top-blowing converting furnace is dynamically adjusted through an execution mechanism. The system comprises a data processing module, a height output module and a feedback loop module. According to the method, the height of the spray gun is continuously corrected and adjusted according to actual conditions, so that the stability and optimization effect of the process are kept.

The invention discloses a household energy storage dynamic optimization method and system combined with load feature learning, which are used for reducing prediction errors caused by sudden loads and improving response efficiency and economical efficiency of a household energy storage system. The method comprises the following steps: calculating a prediction error between real-time total load data and a basic prediction result, identifying and locking one or more high-power electric appliances causing the prediction error as target electric appliances, and decomposing and extracting historical operation state data of the target electric appliances from the total load data; constructing a time sequence feature vector based on the historical operation state data, and training by using a recurrent neural network model to extract the sudden load feature of the target electric appliance; fusing the optimized sudden load characteristics with a basic prediction result to generate a total load prediction curve; and based on the total load prediction curve, combining the residual capacity and the operation health state of the household energy storage system, and adopting a multi-objective optimization method to dynamically generate a charging and discharging strategy.

The invention relates to the technical field of component strength detection, and discloses a big data-based component strength judgment system, which comprises a data acquisition module, a data preprocessing module, a single-point strength analysis module and a continuous strength evaluation module. The data acquisition module obtains multi-dimensional parameters of the component and performs standardization processing, and main body features and abnormal fluctuation areas are separated; the preprocessing module identifies the boundary and calibrates a parameter data space coordinate system; the single-point strength analysis module generates a topological network through key nodes, and judges the strength of a detection unit in combination with structural feature density and span; and the continuous strength evaluation module generates a continuous fluctuation curve based on the parameter sequence, and evaluates the strength continuity through the curvature change rate and the recurrent neural network model. The system realizes multi-dimensional data intelligent analysis, improves the comprehensiveness and accuracy of component strength detection, is suitable for complex structure strength evaluation, and provides technical support for engineering safety.

The invention discloses a low-sample neural network structure reliability evaluation system and evaluation method, and relates to the technical field of engineering structure safety monitoring, and the evaluation system comprises a cloud server which is used for constructing a recurrent neural network model containing a static variable embedding mechanism, completing model training and converting a model format; the edge calculation terminal is used for receiving and preprocessing real-time data of the sensor, executing multi-step prediction to output a future time period response sequence, and calculating a future failure probability through virtual Monte Carlo simulation; the sensor assembly is used for collecting structure state time sequence data; and the communication module is used for realizing data interaction and alarm signal transmission operation. According to the method, collaborative modeling of time-varying and static uncertainty is realized by adopting a static variable embedded recurrent neural network model, failure probability distribution is generated at an edge computing terminal in combination with a virtual Monte Carlo technology, failure risk prediction in a future time period is supported, and real-time and accurate reliability early warning can be realized in a resource limited scene.

The invention provides a method for determining the influence of a straw returning mode on tobacco-rice rotation soil carbon emission, and belongs to the technical field of soil carbon emission. The method comprises the steps that test plots are divided, and different straw returning treatment groups are arranged; installing a static gas tank measuring system and regularly collecting a gas sample; measuring the emission of CO2 and N2O and the concentration of CH4 by using a gas chromatographic method, and calculating the carbon emission flux of the soil; analyzing a straw decomposition process by applying a soil microorganism kinetic equation; analyzing the dynamic change of a soil carbon library based on a carbon recurrent neural network model; constructing a carbon flow network topological structure by using a minimum spanning tree algorithm and identifying key nodes; according to the method, comprehensive influences of environmental factors, crop growth parameters and microbial community structures on carbon emission are integrated through a multi-head cross attention mechanism, accurate quantification and prediction of straw returning carbon emission are achieved, and a scientific basis is provided for optimizing a farmland carbon management strategy and promoting conversion from a farmland carbon source to a carbon sink.

The invention provides a self-adaptive source-load-storage coordinated optimization control method and system, and the method comprises the steps: S1, collecting multi-element load data generated during the operation of a microgrid, and forming a time series data set; s2, key variable characteristics which have the largest influence on the micro-grid load operation state in the micro-grid load requirements are recognized; s3, determining and constructing a source-load-storage cooperative control model based on the key variable characteristics, and determining an optimal decision variable value meeting a real-time load demand; and S4, making a regulation and control strategy according to the optimal decision variable value, and realizing optimal operation of the micro-grid. According to the method, the load demand of the power system is predicted based on the trained recurrent neural network model, the most critical characteristic variable is identified, the time sequence data which can most represent the power grid load operation state during the microgrid operation period is extracted and analyzed, and the unnecessary calculation amount is reduced by focusing on the most critical characteristic variable, so that the power grid load operation state is optimized. Therefore, the whole analysis process is more efficient and accurate.