113 results about "Self-organizing map" patented technology

The present invention relates to the technical field of monitoring, in particular to a remote monitoring method and system for an aviation obstruction light. The method comprises the following steps: on the basis of an external sensor, acquiring electromagnetic signals sent by an aviation obstruction light; and by means of using a signal processing algorithm, processing the obtained original signals to eliminate noise interference and standardize the signal format, so as to generate signal-purified data. Using a support vector machine and a random forest algorithm in the present invention enhances the fault mode identification capability and the accuracy of predicting device performance degradation trends, and substantially improves the reliability of fault prediction; the combination of a Kalman filter and a multi-level decision tree provides powerful support for the integration and analysis of multi-source data, thereby ensuring the comprehensiveness and effectiveness of decision-making support information; and using a self-organizing map network and U matrix visualization technology not only shows advantages in the aspects of data mode identification and anomaly detection, but also improves the interpretability of data analysis by means of visual image displaying.

A method is provided for enhanced data routing in neural networks using Self-Organizing Maps (SOM) integrated with Autoencoder-GAN. The method comprises training an autoencoder to encode input data into a latent space representation; applying a Self-Organizing Map (SOM) to organize the latent space representation into a topological map; refining the latent space representation using a Generative Adversarial Network (GAN), wherein the generator generates enhanced latent space representations and the discriminator evaluates their quality; using the refined latent space representations to update the SOM topology dynamically; generating routing coefficients based on the updated SOM topology to guide data routing in a capsule network; and dynamically adjusting routing within the capsule network using the generated routing coefficients to enhance performance based on the refined latent representations.

The invention discloses an enterprise data asset intelligent analysis method and system based on AI recognition, and the method comprises the steps: receiving an enterprise multi-source heterogeneous data stream, carrying out the joint feature extraction and semantic alignment through a pre-trained multi-modal fusion recognition model, and generating a structured data asset recognition result; constructing a dynamic enterprise data asset atlas according to the structured data asset identification result in combination with the data access trajectory and authority metadata collected in real time; performing spatio-temporal evolution analysis on the dynamic enterprise data asset map, and extracting potential data value density features and risk exposure features; inputting the data value density features and the risk exposure features into a self-organizing mapping network to generate a data asset grading topological graph; and based on the data asset grading topological graph, through strategy constraint reinforcement learning, generating an executable data governance action sequence. According to the embodiment of the invention, the identification precision and real-time analysis capability of special assets of enterprises can be improved.

The invention discloses a dynamic DEM spatial interpolation method which comprises the following steps: performing depression filling, flow direction analysis and confluence cumulant calculation on DEM data, and extracting a natural sub-basin unit by adopting a minimum catchment area threshold method; constructing a topographic feature matrix, performing refined second-level classification on the first-level drainage basin through an improved self-organizing mapping network, generating a hydrological response unit through boundary processing, and establishing a hydrological attribute library; fusing multi-source data, supplementing attribute interpolation such as underlying surface and rainfall, and constructing an interpolation auxiliary parameter system; a drainage basin is divided into regular grids as neurons, a dynamic neural network containing dynamic states and static attributes is constructed, and nonlinear mapping of DEM correction parameters is achieved through optimization of a dynamic activation function and a loss function. The method overcomes the defects that a traditional interpolation algorithm does not consider hydrological boundary constraints, a neural network model topological structure is fixed and the like, and the DEM interpolation precision and the hydrological simulation effect of the complex terrain area are improved.

The invention discloses an unmanned aerial vehicle three-dimensional flight path planning method for wireless sensor network data collection and energy supplementation, and belongs to the technical field of Internet of Things. A network is divided into a plurality of clusters through a density peak clustering algorithm, and a dynamic cluster head selection mechanism based on a routing protocol is designed to optimize the positions of cluster heads. The shortest flight path is determined by adopting a self-organizing mapping network introducing a penetration mechanism. Through a depth deterministic strategy gradient algorithm, the flight height of the unmanned aerial vehicle is dynamically adjusted, and meanwhile, the flight route of the unmanned aerial vehicle is planned, so that the unmanned aerial vehicle can efficiently collect data and supplement energy to sensor nodes in a charging range. According to the method, the distance of the unmanned aerial vehicle for charging the sensor and collecting data is obviously optimized, the energy utilization efficiency is improved, the energy supplement and data collection efficiency of the network is improved, and the network survival time is prolonged.

The invention provides a medical image deep learning auxiliary diagnosis system. The system comprises an image preprocessing module; the feature extraction module is connected with the image preprocessing module; the target detection / segmentation module is connected with the feature extraction module; the classification diagnosis module is connected with the feature extraction module and the target detection / segmentation module; and the result fusion module is connected with the target detection / segmentation module and the classification diagnosis module. According to the invention, a deep learning algorithm, self-organizing mapping, a mixed density network and a clustering algorithm are deeply fused in a feature extraction module, and feature extraction and analysis are carried out on an image from multiple angles. The algorithms cooperate and complement each other, and the limitation of a single algorithm is effectively avoided, so that the system can extract lesion features more comprehensively and accurately, and the disease diagnosis accuracy is improved.

The embodiment of the invention provides a landslide susceptibility prediction method, system and device and a computer readable storage medium. The method comprises the steps of obtaining a to-be-divided topographic image; wherein the topographic image is a slope direction and mountain shadow map extracted based on the digital topographic image; performing region division on the topographic image through a multi-scale algorithm to obtain slope units; wherein the multi-scale algorithm is used for carrying out slope unit division on the terrain; based on the slope unit, environment factor extraction is carried out, and multiple evaluation factors are determined; performing landslide susceptibility prediction on the slope unit based on the plurality of evaluation factors through a landslide susceptibility prediction model to obtain a landslide susceptibility prediction result corresponding to the slope unit; wherein the landslide susceptibility prediction model comprises a self-organizing mapping neural network module, an information amount module and a support vector machine module; the landslide susceptibility prediction model is a model which is trained by adopting historical landslide data and is used for carrying out landslide prediction on the slope unit. According to the scheme, the accuracy of landslide susceptibility prediction can be improved.

The invention provides a coastal erosion rate prediction method, and belongs to the technical field of coastal erosion, and the method comprises the steps: building a three-dimensional seabed grid model, solving a hydrodynamic field through depth average simplification and GPU parallel calculation, and calculating sediment flux distribution based on shear stress discrimination and a high-order windward format. Outputting an erosion mode category and a local erosion strength coefficient by using a coastline erosion feature recognition model comprising a Josephh ring screening layer and a self-organizing mapping projection layer, starting adaptive grid encryption when the local erosion strength exceeds a threshold value, and calling a corresponding parameter set according to an erosion mode to calculate a seabed elevation change and a coastline erosion rate; and a wave energy spectrum reconstruction algorithm is adopted to generate future wave sequence cycle prediction, so that the technical problem that calculation precision and calculation efficiency are difficult to consider in coast erosion rate prediction under a complex wave power condition is solved.

The invention provides a power grid constant value detection method and system, equipment and a storage medium, and belongs to the technical field of power grid detection, and the method comprises the steps: obtaining to-be-detected test constant value data; determining a target neuron matched with the test constant value data from the self-organizing mapping model, and determining a test error between the test constant value data and a weight vector of the target neuron; based on the test error and an error threshold value, determining whether the test constant value data is abnormal or not; wherein the error threshold value is determined by the following steps: acquiring each historical constant value data of the power grid; determining first neurons respectively matched with the historical constant value data from a self-organizing mapping model; and determining an error threshold value based on the weight vector corresponding to each first neuron and the historical constant value data corresponding to each first neuron. According to the power grid constant value detection method and system, the equipment and the storage medium provided by the invention, the recognition precision of the abnormal constant value can be improved.

The invention relates to the technical field of metal spectrum analysis, in particular to a method for determining the element content of a metal sample by using a direct-reading spectrometer. The method comprises the following steps: acquiring a spectrogram of a to-be-analyzed metal sample; further carrying out iterative decomposition on each spectrum peak to obtain a decomposition peak; further inputting the spectrogram and all decomposition peaks of the spectrogram into a spectrum recognition model obtained through self-organizing mapping network training, and recognizing corresponding element categories; and finally, substituting the spectral intensity of each decomposition peak into a pre-established quantitative correction curve of the corresponding element, and calculating to obtain the content of each element in the metal sample without depending on a priori standard spectrogram or a preset overlapping peak number, so that accurate element quantitative determination under a complex overlapping spectrum peak is realized, and the determination precision and adaptability are improved.

The invention discloses a tree line discharge fault identification method based on big data fault traveling wave current characteristics, and the method comprises the steps: S1, collecting fault traveling wave current waveform data and environment data, and enabling the environment data to comprise wind speed, humidity and temperature data; s2, performing adaptive discrete wavelet transform on the fault traveling wave current waveform to obtain a fault wavelet coefficient, obtaining an enhanced fault wavelet coefficient through an improved dynamic attention network based on the fault wavelet coefficient, and obtaining the optimal frequency band energy of the enhanced fault wavelet coefficient; splicing the enhanced fault wavelet coefficient, the optimal frequency band energy and the environment data into a comprehensive feature vector; s3, introducing a self-organizing mapping network based on the comprehensive feature vector to construct an optimized state space, collecting traveling wave current waveform data in real time, and obtaining waveform time sequence similarity between the traveling wave current waveform data and the optimized state space by using an improved dynamic time warping algorithm; when real-time data is input, subspace matching of corresponding environmental conditions can be quickly positioned.

The invention belongs to the field of control, relates to a resin production waste gas recovery control method and system, and aims to solve the technical problems that in the prior art, working condition division cannot be matched with the real state of a multivariable system, and PID parameters are balanced among a plurality of control targets. The neurons are divided into stable working condition neurons, transition working condition neurons and disturbance working condition neurons; s2, obtaining a Pareto optimal solution set composed of a PID parameter group, wherein the Pareto optimal solution set comprises an ITAE optimal solution, a control output change rate optimal solution and an equilibrium solution; s3, determining an optimal matching unit of the potential feature vector in the self-organizing mapping network; s4, when the optimal matching unit is a stable working condition neuron and the absolute value of the current control error is greater than a high-order threshold value, selecting an ITAE optimal solution; when the absolute value of the current control error is smaller than a low-level threshold value, the optimal solution of the control output change rate is selected; and a balance solution is selected between the two. Accurate identification and division of the operation state of the resin production waste gas recovery process are realized.

The invention discloses an intelligent segmentation and feature analysis method for a medical image, and relates to the technical field of image analysis, and the method comprises the steps: reading a multi-modal medical image, and carrying out the preprocessing and size standardization; performing medical image segmentation based on a U-Net structure, performing edge detection by adopting a Sobel operator, and optimizing the edge of the segmented image in combination with morphological operation to obtain an edge-optimized focus segmentation map; feature point matching is carried out on the multi-modal focus segmentation image based on self-organizing mapping, and non-rigid registration is carried out through TPS transformation based on matched feature points; carrying out foreground region enhancement operation on the registered image, introducing a channel attention module based on a DenseNet-201 model to carry out feature extraction on the registered image, and optimizing features by using a whale optimization algorithm to generate an optimal feature subset; and carrying out illness state classification on the feature subsets by an adaptive neural fuzzy inference system optimized based on a genetic algorithm.

The invention discloses a harmonic current signal-based high-voltage cable defect identification method and system. The method comprises the steps of obtaining harmonic current monitoring data of a high-voltage cable metal sheath; analyzing the harmonic current monitoring data by using a principal component analysis method to obtain a principal component characteristic quantity; determining a defect identification initial model based on a self-organizing mapping algorithm, and training and optimizing the defect identification initial model based on the principal component characteristic quantity to obtain a defect identification optimal model; and inputting harmonic data of a to-be-detected high-voltage cable into the defect identification optimal model so as to identify the defect type of the to-be-detected high-voltage cable. Amplitude-frequency characteristics of harmonic current signals are extracted through principal component analysis, abnormal harmonic data generated by different defects are classified and distinguished through a self-organizing mapping algorithm, and therefore defect identification based on high-voltage cable line harmonic signal abnormity is achieved. And reliable technical support can be provided for cable defect identification in a power system in practical application.

The invention discloses an electric pile state identification and classification method, equipment and a storage medium, and relates to the field of battery data calibration. The method comprises the following steps: according to the similarity between each type of galvanic pile parameters and galvanic pile voltage, determining the correlation between each type of galvanic pile parameters and galvanic pile voltage; selecting galvanic pile parameters with large correlation as training data; training the training data through a self-organizing mapping neural network to obtain an optimal clustering model for clustering the pile parameters; clustering the training data through the optimal clustering model to obtain a plurality of clusters of clustering data; generating a pseudo pile state label for each type of clustering data and then forming a sample set; and training the sample set through a one-dimensional convolutional neural network to obtain a pile state classification model for pile state classification. According to the method, the workload of manual labeling is reduced, and the state of the galvanic pile can be effectively identified and classified.

This application belongs to the field of control and relates to a control method and system for recovering waste gas from resin production. It addresses the technical problems in existing technologies, such as the inability of operating condition division to match the true state of a multivariable system and the need for PID parameters to balance multiple control objectives. The control method includes the following steps: S1, dividing neurons into stable operating condition neurons, transitional operating condition neurons, and disturbance operating condition neurons; S2, obtaining a Pareto optimal solution set composed of PID parameter groups, including the ITAE optimal solution, the control output rate of change optimal solution, and the equilibrium solution; S3, determining the best matching unit for the latent eigenvectors in a self-organizing map network; S4, when the best matching unit is a stable operating condition neuron, selecting the ITAE optimal solution when the absolute value of the current control error is greater than a high threshold; selecting the control output rate of change optimal solution when the absolute value of the current control error is less than a low threshold; and selecting the equilibrium solution when it is between the two. This achieves accurate identification and division of the operating state of the resin production waste gas recovery process.

A voice generation method and apparatus, a product, a device, and a medium, which are applied to the technical field of voice generation. The method comprises: using a quantizer to discretize a voice feature vector of an original voice signal, obtaining a discrete symbol representation corresponding to the original voice signal (S11); extracting a text feature corresponding to a target text (S12); and inputting the text feature and the discrete symbol representation into a voice generation model, so that the voice generation model uses the text feature as a condition, and generates a target voice on the basis of the discrete symbol representation (S13). The quantizer is obtained by training a self-organizing map network by means of voice signal training samples. The method may restore an original voice feature more accurately, and improve the quality of the generated voice.

This invention discloses a random flood forecasting method based on machine learning and cloud models. The steps are as follows: using typical historical floods as basic data for model calibration and testing, selecting classification indicators that meet the conditions, and classifying historical floods based on a self-organizing map neural network (SOM); using the maximum information coefficient method (MIC) to screen the impact factors of classified floods, and establishing a random flood forecasting model based on different machine learning methods; for different types of floods, solving the fusion weights of different forecasting models based on the cloud model, and weighting the simulation results of each model to obtain the model integrated forecast result; analyzing and calculating the relative forecast error, and establishing the joint distribution function of the relative forecast errors at adjacent moments based on the Copula method; finally, obtaining real-time online flood information to achieve random flood forecasting. This invention can improve the accuracy of flood forecasting and provide a new approach to hydrological prediction and forecasting.

The invention relates to a commercial vehicle accident high-risk scene portraying method and system. The method comprises the following steps: acquiring original data; constructing a high-risk scene index system, and defining a risk level; preprocessing the original data based on the high-risk scene index system to obtain preprocessed data; training the self-organizing mapping neural network by utilizing the preprocessed data, and adjusting network node weights by adopting a forgetting suboptimal rule to obtain an initial clustering center; and taking the initial clustering center as a starting point, taking the risk level as an initial K value, and carrying out K-means clustering analysis on the preprocessed data to obtain a commercial vehicle accident high-risk scene portrait result. According to the method, the SOM neural network and the weight updating mechanism with the forgetting factor are introduced to optimize the clustering effect, and the pendulum effect generated in the later period is avoided; through methods of clustering analysis, multiple correspondence analysis and the like, accurate description of an accident high-incidence scene of the commercial vehicle is realized, and a support is provided for comprehensively improving the safety of the road transport vehicle.

A system and method for training relational networks. A method includes applying a self-organizing map (SOM) to training data in order to create a visualization. The SOM is a neural network configured to transform relationships between data items. The visualization has a lower dimensionality than the training data. The method also includes training machine learning models of a generative relational network (GRN) based on the visualization, where the GRN includes sets of nodes having respective machine learning models among the machine learning models of the GRN and the sets of nodes include a set of dominance factor nodes and a set of evolution of internal component nodes. The set of dominance factor nodes defines a dominance factor based on change intensity and change frequency, and the set of evolution of internal component nodes defines evolution with respect to changes determined based on values of the dominance factor over time.

The present invention provides a method and system for classifying loads of high-energy-consuming industrial users, electronic equipment, and media, which are applied to the field of intelligent power consumption technology. The method includes: considering the load variation characteristics, load utilization rate, and user's sensitivity to electricity prices, determining multi-dimensional load characteristic indicators based on the daily load data of each industrial user; extracting time characteristics of the daily load data of each industrial user based on a convolutional autoencoder, and determining the power consumption characteristics based on the extracted load time characteristics and multi-dimensional load characteristic indicators of each industrial user; inputting the power consumption characteristics into a self-organizing map neural network to determine the output node to which each input data is mapped; screening out the target output node based on the number of input data corresponding to each output node; and clustering the power consumption characteristics of each industrial user using the target output node as the initial clustering center. The present invention solves the problems of poor clustering effect, large computational complexity, and low processing efficiency in the industrial load clustering process.

The invention discloses a supply chain multi-modal interactive question and answer method based on a large language model, and the method specifically comprises the steps: S1, collecting text, voice, image and table data in a supply chain scene, and extracting semantic features to form a multi-modal semantic vector; s2, extracting inventory, transportation, production and order time sequence data, and inputting the data into the improved self-organizing mapping neural network to generate a semantic state topological structure; s3, executing concept drift detection and locally reconstructing nodes, and outputting a stable supply chain semantic state vector; s4, fusing the supply chain semantic state vector and the user context information to generate context state enhanced representation; s5, calculating a multi-modal correlation weight to realize semantic alignment and unified coding; and S6, inputting the large language model and combining with knowledge graph reasoning to generate text, voice or chart answers. According to the method, supply chain multi-modal information intelligent fusion and semantic question and answer accurate generation are realized, and the decision-making efficiency and the intelligent interaction level are remarkably improved.

The invention discloses a 5G communication modulation signal identification method and system, and belongs to the technical field of signal identification, and the 5G communication modulation signal identification method comprises the steps of signal receiving, spatial feature separation, multi-domain feature extraction and feature fusion and identification. According to the invention, spatial spectrum estimation is carried out on the multipath beam superposition signals received by the MassiveMIMO array by using the MUSIC algorithm, efficient separation of N paths of independent beam signals is realized, and the technical problem that a traditional single-antenna model cannot adapt to a multi-antenna scene is thoroughly solved; according to the method, the sparse Transform model and the stacked self-organizing mapping are combined, accurate classification and proportion calculation of QPSK, 16QAM, 64QAM and 256QAM hybrid modulation signals are achieved, the recognition accuracy is larger than or equal to 92% when SNR = 0dB, the recognition accuracy is improved by more than 18% compared with a traditional algorithm, and the problems that high-order modulation constellation points are fuzzy and hybrid modulation classification is difficult are effectively solved.

The invention provides a renewable resource intelligent recovery data management system and method based on the Internet of Things, and relates to the technical field of renewable resource intelligent recovery. The method comprises the following steps: collecting original putting data of resident recycled materials in real time; performing preliminary cleaning and structured processing on the original delivery data by using an edge computing device to obtain cleaned delivery data; uploading the cleaning delivery data to a cloud data platform through a communication protocol; constructing a neural ODE network model, and predicting the resource recovery amount of each region in n days in the future; the cloud data platform performs tagging classification on user recovery behaviors based on self-organizing mapping analysis; and encrypting and storing the cleaning and putting data in the transmission and analysis process by adopting an encryption algorithm. The neural ODE network model and the self-organizing mapping analysis expand the application scene of the data, improve the environmental protection atmosphere and the resource recovery efficiency in the region, deeply mine the data value, and create more development opportunities for the resource recovery service.

The invention discloses a method and a system for monitoring internal discharge abnormity of a switch cabinet, and aims to realize accurate monitoring and visual display of internal discharge of an all-insulated closed inflatable high-voltage switch cabinet by combining an ultraviolet image transmission fiber bundle and a big data analysis technology, so that the online monitoring capability of an equipment operation state is improved, and the operation safety of the switch cabinet is improved. And safe operation of power equipment is ensured. According to the system, ultraviolet light signals in a switch cabinet are acquired by using ultraviolet image transmission fiber bundles, photoelectric conversion and signal amplification are performed through an APD avalanche photoelectric detection module, and historical data modeling, feature extraction and clustering, real-time data monitoring and anomaly detection and classification are performed in combination with an autoregression model, a self-organizing mapping neural network and a density clustering algorithm. The system can realize real-time monitoring and abnormity feedback of the equipment state in severe environments such as strong electromagnetic interference, high temperature and high pressure, provides high sensitivity and anti-interference capability, and ensures monitoring accuracy and timeliness.

Systems, methods, and apparatuses for performing real-time cytometry data analysis. One apparatus includes at least one electronic processor and at least one memory storing instructions executable by the at least one electronic processor. The at least one electronic processor is configured, through execution of the instructions, to obtain flow cytometry data generated by a cytometry instrument representing cells of multiple categories, generate a feature vector representation based on the flow cytometry data using a plurality of self-organizing maps (SOMs), wherein each SOM corresponds to a different category of multiple categories, and predict each of one or more target labels of the cells by applying each of one or more regression models to the feature vector representation.

The embodiment of the application provides an image clustering method and an electronic device, the method comprises the following steps: determining a plurality of images, and determining the structural similarity between the images to generate a structural similarity matrix. The structural similarity matrix is input into a self-organizing mapping network, and the self-organizing mapping network finds the best matching neuron node for a first image included in each image pair based on a competitive learning rule according to the structural similarity matrix, and maps a second image included in each image pair and having structural similarity with the first image to a topologically adjacent neuron node to generate a mapping result of the plurality of image pairs. Based on the mapping result, a multi-level clustering process is performed to obtain a fine clustering result. In this way, based on the combination of structural similarity and the self-organizing mapping network, visual-sensitive similarity measurement and hierarchical clustering division are realized, the clustering fineness is improved, and the operation and maintenance cost is reduced.

The invention relates to a gamma photon positioning method and device for an orthogonal strip-shaped tellurium-zinc-cadmium detector. The method comprises the following steps: acquiring current signals respectively generated by interaction of a plurality of gamma rays and the orthogonal strip-shaped tellurium-zinc-cadmium detector; vectorizing the plurality of current signals to obtain an input vector set; an initialized neural network is constructed based on the positions of a plurality of electrode strips in the orthogonal strip-shaped cadmium zinc telluride detector, and the number of initialized neurons is the same as that of the plurality of electrode strips; performing weight vectorization processing on the plurality of initialized neurons to construct an initialized weight vector set; and based on the input vector set and the initialized weight vector set, iteratively reducing the range of the initialized weight vector set through a self-organizing mapping algorithm to obtain a target weight vector, and determining the position of the gamma photon. The purpose of determining the position of the gamma photon based on the neural network and the self-organizing mapping algorithm is achieved, and therefore the technical effects of improving the accuracy of the position of the gamma photon and improving the position obtaining efficiency are achieved.

The invention relates to the technical field of new energy power prediction of an electric power system, and discloses a mountain area power distribution network distributed photovoltaic prediction method, system and device considering prediction errors and a medium, and the method comprises the steps: firstly collecting meteorological data, historical output data and topographic data of a distributed photovoltaic field station with a meteorological observation capability; based on multi-source data, clustering stations by adopting a dynamic time warping and dynamic self-organizing mapping algorithm, and dividing mountainous climate similar regions; constructing a photovoltaic output prediction model fusing a long short-term memory network and a time convolution network for each region, and generating a preliminary prediction value; on the basis, establishing a conditional error probability distribution model by utilizing a prediction error sequence; and finally, all stations are corrected and predicted by combining the prediction model and the error model, and a cluster output result is output. The method effectively improves the prediction precision and reliability in a mountainous area high-proportion distributed photovoltaic access scene, and is especially suitable for a station lacking local meteorological data.

The invention discloses a power distribution network load prediction method and system considering user behavior characteristics, and the method specifically comprises the steps: firstly building a multi-dimensional prediction index system based on the user behavior characteristics based on the historical load operation data and natural environment meteorological data of a power distribution network; then, based on an index system, clustering analysis is carried out on historical load data through K-mean value and self-organizing mapping algorithm fusion, and different load features are extracted; then, adaptive optimization is carried out on model parameters of a time convolutional network (TCN) by adopting a myxobacteria optimization algorithm, and multi-type loads are predicted based on a TCN hyper-parameter optimization result; and finally, comparing and analyzing the prediction result by using the evaluation index. According to the method, the load prediction precision of the power distribution network is improved, the problems of active power uncertainty and voltage fluctuation caused by photovoltaic power generation are reduced, and the overall planning, fine scheduling and safe and stable operation capabilities of the power distribution network are improved.