3130 results about "Encoder decoder" patented technology

The invention provides a chip production logistics optimization scheduling method and system based on deep reinforcement learning, and relates to the technical field of deep reinforcement learning. An improved bidirectional long-short-term memory network and a graph neural network are used for extracting a material flow time sequence rule, equipment operation cycle characteristics and logistics network topology characteristics, a fusion state representation is generated through a contrast learning network, a layered and collaborative deep reinforcement learning architecture is constructed, an improved encoder-decoder structure is used for generating a strategy vector, and the strategy vector is constructed. And performing hierarchical joint training through a multi-agent execution network, inputting a real-time state and a strategy vector into the trained deep reinforcement learning architecture, generating a scheduling instruction, and performing hierarchical model prediction control in combination with a Bayesian neural network ensemble model and a recurrent neural network prediction model to realize optimized scheduling of chip production logistics.

The invention discloses a photovoltaic power station unmanned inspection method and system based on multi-mode fusion detection, and relates to the technical field of intelligent operation and maintenance of photovoltaic power stations. The method mainly comprises the following steps: (1) collecting multi-modal data of a photovoltaic module and carrying out preprocessing operation; (2) uniformly mapping the heterogeneous data into discrete marks by using each pre-trained modal data marker; (3) training a multi-modal fusion detection network based on a Transform encoder-decoder architecture by using the training sample, and adjusting network parameters to obtain a trained model; and (4) outputting a defect type, a position bounding box and a severity score by utilizing the trained model according to the obtained multi-modal data or single-modal data of the photovoltaic module. The problems that traditional single-mode detection is high in omission ratio and multi-mode fusion is low in efficiency are solved, the inspection collection module, the data processing module, the multi-mode fusion detection module and the fault decision module are provided in a targeted mode, and full-process automation from data collection to intelligent decision is achieved. The operation and maintenance cost of the photovoltaic power station is obviously reduced; and the fault response efficiency is improved.

The invention discloses a lightweight semantic segmentation network and method for power grid inspection, and the method comprises the steps: employing an encoder-decoder architecture, inputting original data into a lightweight MobileNet v2 backbone network for feature extraction at an encoder part, inputting an obtained feature map into a lightweight large-scale ASPP module, and carrying out the feature extraction; further performing multi-scale feature extraction on the high-dimensional feature map; and in the decoder part, high-dimensional features and low-dimensional features obtained by the backbone network are respectively input into a cross attention wavelet fusion module, after an obtained feature map is fused with the output of a decoder, the feature map is recovered to the size of original data through down-sampling operation by a full-dimensional dynamic convolution post-processing module, and a final prediction mask is obtained. According to the invention, by optimizing a coding and decoding structure, the receptive field of the model and the modeling capability of a narrow and long target are improved, the balance between the precision and the complexity can be effectively realized, and the method is suitable for an unmanned aerial vehicle inspection system with limited resources.

The invention relates to an underwater image enhancement method based on frequency domain analysis and visual Mama, comprising the following steps: constructing a multistage wavelet transform network, the network being an encoder-decoder structure, the encoder comprising a space-frequency feature fusion module and discrete wavelet transform; the decoder comprises a space-frequency feature fusion module and inverse discrete wavelet transform; an underwater image is decomposed into a low-frequency part and a high-frequency part through discrete wavelet transform in an encoder, spatial feature extraction and frequency feature extraction are conducted on the low-frequency part and the high-frequency part through corresponding space-frequency feature fusion modules based on Mama and frequency domain convolution respectively, and splicing is conducted on the low-frequency part and the high-frequency part along a channel in the decoder according to the sequence of low frequency and high frequency. Reconstruction is carried out through inverse discrete wavelet transform; and inputting the underwater image data into the multistage wavelet transform network to obtain enhanced underwater image data. Compared with the prior art, the method has the advantages of improving the enhancement quality, reducing the complexity, quickly enhancing and the like.

The invention discloses a vehicle intelligent driving decision optimization method based on deep learning. The method belongs to a vehicle intelligent driving decision technology. Firstly, various sensors are installed on a vehicle to collect data, and data sets are divided according to a proportion after cleaning and marking. A model containing an encoder-decoder and an attention mechanism is constructed, the weight is initialized, and a hyper-parameter and anti-overfitting strategy is set. And training the model by using a specific algorithm, calculating a loss value and evaluating and adjusting according to a verification set. After the model is integrated to the system, a preliminary candidate decision is determined according to a probability threshold, the decision is optimized in combination with a vehicle surrounding environment information set and a risk assessment function, and decision probabilities are redistributed in real time according to environment changes by means of an environment perception feedback mechanism. The method overcomes the defects of the traditional technology, utilizes the advantages of deep learning, effectively improves the driving decision precision, enhances the adaptability of the system to complex road conditions and environments, and guarantees the safe and efficient operation of intelligent driving.

The invention discloses a person and post matching method, system and equipment based on Transform and a medium, belongs to the technical field of cross application of natural language processing and human resource management, and aims to solve the technical problem of how to reduce subjective errors in a person and post matching process, improve matching accuracy, fairness and objectivity and improve manual matching efficiency at the same time. According to the technical scheme, the method comprises the following steps: data collection: collecting post information data and job seeker resume data from a plurality of data sources, and cleaning and sorting the collected data to obtain a post description text and a resume text; text preprocessing: preprocessing the acquired and sorted post description text and resume text to obtain the preprocessed post description text and resume text; the method comprises the following steps of: fusing a Transform model and a BERT model; constructing an encoder-decoder structure based on the Transform model; optimizing the model; multi-label text classification: constructing a multi-label classification system; and outputting and evaluating a matching result.

A method for robot path planning using skills extracted from human-taught motion programs applied to a new obstacle environment. A three-dimensional convolutional neural network is used to extract features characterizing an obstacle environment, where the feature vector representation of the obstacles overcomes problems encountered when using point cloud obstacle data. The obstacle feature data and robot path start and goal points are provided to an encoder / decoder neural network system which is trained to extract skills from a database of human-generated motion programs. The encoder / decoder neural network system produces a distribution of waypoints for the current obstacle environment and start / goal points. The distribution of waypoints is used to perform a final collision-free path generation using either a rapidly-exploring random tree (RRT) technique or an optimization-based technique.

The invention provides a microcrack precise segmentation method integrating feature fusion and convolution attention, and belongs to the field of image processing. According to the method, a crack segmentation network based on an encoder-decoder architecture is constructed, a convolution block attention module is introduced at an encoder end, background noise is adaptively suppressed and obvious characteristics of cracks are enhanced through a channel and space dual attention mechanism, and the method is suitable for the adaptive segmentation of the cracks on the premise of almost not increasing the calculation overhead. The sensitivity of the model to microcracks is improved; a feature fusion module is introduced at a decoder end, and cooperation of low-layer details and high-layer semantics is realized through cross-layer fusion, so that a semantic gap is effectively bridged, detail loss caused by traditional convolution stacking is avoided, and continuity and a complete topological structure of a long and narrow crack are ensured. According to the method, through collaborative optimization of multi-scale feature extraction and an attention mechanism, accurate capture of the saliency features of the crack and effective suppression of complex background interference are realized, and the detection sensitivity and overall segmentation consistency of the micro-crack are remarkably improved.

A method and apparatus for reconstructing a three-dimensional (3D) scene from a two-dimensional (2D) input image of the scene using a fully-differentiable transformer-based encoder-decode. A 2D input image encoded into a set of image features using a pre-trained vision transformer model, wherein the vision transformer model is pre-trained with multi-view RGB image supervision and point cloud supervision. The set of image features is projected onto a 3D triplane representation using a transformer decoder to obtain output triplane tokens. A triplane representation is created from the tokens and queried. 3D point features of color and density for volumetric rendering re predicted using a multi-layer perceptron. The geometry of the generated 3D asset is represented with a surface mesh including vertices and triangular faces. A texture map by is created with a multichannel image in UV space. Multiple views of the 3D scene are simultaneously generated based on the surface mesh.

The invention provides a three-dimensional seismic fault identification method based on double-attention multi-scale fusion U-Net. The method comprises the following specific steps: constructing a U-shaped network architecture comprising an encoder, a decoder and jump connection; a constructed multi-scale feature fusion module is embedded in the first level of the encoder, and the extraction capability of fault features of different scales is enhanced through a multi-branch structure; introducing the constructed hole fusion modules into the second and third levels of the encoder, designing and expanding a receptive field by using multiple expansion rates, and capturing fault structures of different scales; a double-attention parallel mechanism is integrated in jump connection, and the sensitivity of channel attention and space attention to fault features is improved; constructing a combined loss function; and finally, performing three-dimensional seismic data training and reasoning based on the optimized model to realize high-precision fault identification. The method has high generalization and accuracy, and especially has good performance in the aspect of seismic image fault identification containing a large fault scale span.

The invention discloses a hypergraph fusion-based power Internet of Things anomaly detection method and system. The method comprises the steps of capturing and preprocessing multi-source heterogeneous data in a power Internet of Things; learning potential feature representation of the multi-source heterogeneous data by using an encoder-decoder consisting of an LSTM (Long Short Term Memory) and a Transform encoder, and adding adversarial training in a training process; performing clustering analysis on the learned potential feature representation, and constructing a hypergraph according to an analysis result; carrying out weighted fusion on a plurality of hypergraphs by adopting an HGCN-Transform mode, and learning hypergraph feature representation on the fused hypergraphs by adopting a hypergraph attention mechanism; the output of the hypergraph attention mechanism is predicted through a Transform encoder, reconstruction is achieved through VAE, the abnormal score is calculated through a predicted value and the reconstruction probability, a threshold value is set, the abnormal score is compared with the threshold value, whether the abnormal score is abnormal or not is judged, and anomaly detection is completed. According to the method, complex and diversified relationships among multi-source heterogeneous data can be mined, and the anomaly detection accuracy is improved.

Automatic recognition of non-verbal messages in speech, and in particular to detection or analysis of prosodic multilayered analysis of intonation units such as prosodic unit prototypes and their multi-labeled variations, may form a hierarchical classification for the analysis of non¬ verbal information or cues in speech. A speech captured by a microphone is fed to a weakly- supervised deep learning acoustic model for speech recognition and transcription, that may be based on encoder-decoder Transformer architecture, such as Whisper by OpenAI. The model is trained to output multiple words form the text in the captured speech, to identify Intonation Units (IUs) that include one or more words, and associate non-verbal labels to each of the IUs. The labels may indicate a prototype, a discourse function (such as a conversation action), an emotion, an emphasis, or an attitude, as well as a genre of a part of, or whole of, the entire captured speech.

Disclosed in the present invention is a non-autoregressive Transformer-based modeling method for a 4-level pulse amplitude modulation high-speed transmitter. The method involves establishing a deep learning model having an encoder-decoder architecture to predict the behavior of a 4-level pulse amplitude modulation transmitter. An encoder processes unordered non-sequential inputs, including an input signal parameter and link parameters, to generate a context vector and then transmit same to a decoder. The decoder uses both the context vector generated by the encoder and a transmitter output signal sequence to generate a categorical probability distribution for each point in the sequence one by one. The model is trained using a random masking strategy, and inference is performed by means of non-autoregressive decoding and filtering, so that the model can perform parallel prediction on an output sequence, and perform a filtering process to predict an output signal. Compared to traditional simulation methods, the present invention achieves a significant acceleration effect, particularly when processing multi-link systems.

The invention discloses an infrared small target detection method fusing local prior and a multi-scale global background, and the method comprises the steps: firstly obtaining image data containing an infrared image and a mask label corresponding to the infrared image, and carrying out the preprocessing; secondly, a target detection model of an encoder-decoder architecture is constructed, an encoder comprises a local detail prior mining branch and a multi-scale global background perception branch which are parallel, step-by-step feature extraction is performed on the preprocessed image data, and a decoder comprises a progressive feature fusion decoding branch; and inputting the features of each level of the encoder double branches into decoder branches for decoding step by step to obtain a detection result. And finally, a weighted depth supervision mechanism is introduced in training, auxiliary prediction output is set in a plurality of decoding layers, and weighting loss is calculated. According to the method, the problems of insufficient local detail modeling, insufficient multi-scale global background perception of Mamba, difficulty in global and local feature fusion and the like in the existing method are solved, and the detection precision of the infrared small target is improved.

The invention relates to the technical field of wireless communication, and discloses an intelligent channel prediction method based on a deep multi-layer perceptron architecture in order to solve the problem of high-dimensional near-field cascade channel prediction in a reconfigurable intelligent surface assisted low-orbit satellite communication system. Designing a frequency domain MLP architecture to learn global time sequence characteristics of a cascade channel; secondly, considering high-dimensional spatial domain characteristics of a near-field cascade channel, decomposing the cascade channel into a plurality of feature dimensions, and proposing a multi-branch MLP architecture to realize global spatial domain correlation modeling; and finally, by taking the frequency domain MLP and the multi-branch MLP as basic network feature extraction modules, designing a channel prediction network model based on an encoder-decoder architecture. According to the method disclosed by the invention, the accurate prediction of the downlink CSI can be realized by utilizing the historical channel state information estimated by the uplink communication, and the proposed channel prediction model has better channel prediction precision and model convergence speed.

The invention provides a medical image segmentation method based on wavelet transform and an attention mechanism, and belongs to the technical field of image segmentation of medical image processing. The technical problem that semantic differences between global features and features of coders and decoders at the same level cannot be captured due to limitation of a traditional CNN receptive field is solved. According to the technical scheme, the method comprises the following steps of 1, preprocessing an input polyp medical image; step 2, constructing an encoder decoder network based on wavelet transform and an attention mechanism; 3, inputting the preprocessed polyp image data set into an encoder-decoder network of a wavelet transform and attention mechanism for training, and reserving an optimal model; and step 4, after training is completed, inputting the polyp medical image verification set into the obtained optimal model. And detecting the focus segmentation effect of the model in the polyp medical image verification set. According to the method, the polyp lesion can be segmented more accurately, and the accuracy and robustness are improved.

The invention relates to a medical image segmentation method and system based on residual Mama and multi-scale boundary enhancement. The method comprises the following steps: acquiring and preprocessing a medical image; inputting the image into a segmentation model based on an encoder-decoder architecture; the encoder synchronously extracts local texture features and models long-range spatial dependence through residual error convolution blocks and residual error Mama blocks which are alternately connected; fusing and enhancing the jump connection features between the encoder and the decoder through a boundary enhancement module to optimize boundary characterization; integrating a multi-scale gating attention module in a decoding path, and adaptively selecting and fusing multi-scale context features; and finally outputting the high-precision segmentation mask. The method effectively solves the problems that in the prior art, long-range dependence and local details are difficult to consider, the multi-scale feature fusion capability is insufficient, boundary segmentation is fuzzy and the like, and the segmentation accuracy, the boundary continuity and the clinical practicability are remarkably improved.

The invention discloses an adaptive mask medical image segmentation method based on a self-supervised mask and deep reinforcement learning, and the method comprises the steps: employing a classic encoder-decoder architecture for a self-supervised mask reconstruction network, fusing a Swin Transform encoder, and carrying out the feature fusion of local image blocks through a self-attention mechanism; according to the self-adaptive mask model, a PPO deep reinforcement learning algorithm is adopted, a strategy network and a value network are constructed, mask actions are dynamically regulated and controlled, reconstruction errors are gradually reduced, a mask strategy is continuously optimized in multiple times of strategy updating for self-adaptive optimization, and high-quality reconstruction of a medical image influenced by missing information is achieved; according to the method, high-quality feature representation can be obtained in an unlabeled data environment, and relatively high precision and accuracy are presented on a public data set.

The invention belongs to the technical field of information technology and image recognition, and relates to a handwritten mathematical formula recognition method and system. The method comprises the following steps: constructing handwritten mathematical formula annotation data; constructing an initial handwritten mathematical formula recognition model, training the initial model by using the annotation data, and performing joint optimization on a sequence decoding task and a tree structure prediction task to obtain a handwritten mathematical formula recognition model; and carrying out handwritten mathematical formula identification by adopting the handwritten mathematical formula identification model obtained by training. According to the method, the advantages of a sequence decoding model and a tree decoding model are fused, the understanding and generalization ability of the model to a mathematical formula structure are effectively improved, and the accuracy of handwritten mathematical formula recognition is remarkably improved on the basis of keeping the efficient parallel training advantages of an existing encoder-decoder model-based method.

The invention discloses a complex target recognition method based on a deep twin self-attention network, and relates to the technical field of target recognition methods. According to the method, a twin network based on a sliding window multi-head self-attention transformation mechanism is applied, the global modeling capability of a feature extraction network and the acquisition efficiency of long-distance semantic information are enhanced, and the bottom-layer feature extraction capability and the abstract effect of advanced semantic information are improved; the twin network personalized modeling improves the feature segmentation effect; the whole network forms a U-shaped structure, a local symmetric encoder and decoder structure is adopted, global context information is supplemented, and semantic information space association is enhanced; a residual link is added, so that the deep network is trained efficiently, and the recognition precision is enhanced; and adaptive image feature scale input processing is added, and adaptive scale transformation is performed on the input image, so that an accurate end-to-end semantic segmentation result is realized, and multi-scale input is realized to enhance the generalization ability and accuracy of a network model.

The invention discloses an image deblurring system and method based on Transform and multi-scale feature fusion, and relates to the technical field of image processing, the system comprises an input layer, a processing layer and an output layer, the processing layer comprises a multi-input multi-output encoder-decoder module, the multi-input multi-output encoder-decoder module is used for extracting features from blurred images of different scales through an encoder, and outputting the features to the output layer; a clear image is gradually restored based on the extracted features through a decoder; the multi-feature multi-scale fusion module is used for integrating local and global feature information and enhancing the learning ability of the system to global features; according to the improved Transform module, a self-adaptive window mechanism and a cross-channel attention mechanism are introduced, so that the system can better adapt to different types of blurred images, and the effects of feature extraction and image restoration are improved; and the attention fusion module dynamically learns and integrates useful feature information by introducing a channel attention mechanism. The image deblurring method is suitable for various image deblurring application scenes, and image detail features can be effectively recovered.

The invention discloses a vacuum preloading soil settlement prediction and control method based on deep learning, and aims to realize precise prediction and dynamic closed-loop control of soil settlement. According to the method, an encoder-decoder architecture is adopted to construct a model framework, and off-line generation of a proprietary model is completed through multi-measurement-point space-time joint training and a combinatorial optimization strategy; multi-source data of a target monitoring point is collected and processed to generate a sequence sample of an adaptive model; in order to adapt to the time-varying characteristics of the geological attributes, deploying an incremental online learning mechanism, and solidifying and finely tuning model parameters through an elastic weight; and based on a prediction result of the updated model, a PID control unit calculates a vacuum degree adjusting amount, a vacuum pump control system is driven to dynamically adjust the operation vacuum degree, and it is ensured that soil settlement is within a reasonable range. Finally, a cooperation mechanism of high-precision settlement prediction and dynamic vacuum degree regulation and control is constructed, and the settlement control precision, the response speed and the construction safety are remarkably improved.

The invention discloses a remote sensing image semantic segmentation-oriented spatial spectrum context feature enhancement method and device. The method comprises the following steps of: constructing a neural network model based on an encoder-decoder architecture; inputting the preprocessed remote sensing image set into a feature coding block, and extracting hierarchical features; according to the model, spectrum attention and position self-attention mechanisms are integrated through a collaborative attention module, and the sensitivity of network spectrum and spatial context information is remarkably enhanced; the network coding module efficiently compresses and codes the features before entering a decoding stage; a cross fusion module is applied to intelligently fuse features between a coding block and a decoding block to ensure that information is not lost in a decoding process; the feature map is mapped into a pixel-level category label through a Softmax layer, precise semantic segmentation is achieved, and a high-quality classification result is obtained. According to the method, through collaborative modeling of the spectrum and the spatial context, high-frequency details are effectively reserved, and the precision of semantic segmentation is improved.

The invention discloses a dam break flow rapid prediction method and a dam break flow rapid prediction system adopting a Transform-ResUNet model. Comprising the steps of obtaining actual dam break data or obtaining a data set for model training through numerical simulation; a dam break flow space-time coupling prediction model fusing Transform and ResUNet is constructed, the dam break flow space-time coupling prediction model is an encoder-decoder architecture, an encoder extracts spatial features of a water flow field through a plurality of residual convolutional layers, then feature maps are partitioned and flattened into time sequence features, and the time sequence features are input into a parallel stacked Transform module for time dynamic modeling; the decoder fuses the spatial features and Transform enhanced time sequence features through cross-layer connection, and reconstructs a predicted water flow field at a future moment through bottleneck residual blocks which are up-sampled and stacked step by step; and training the constructed dam break flow space-time coupling prediction model, and realizing rapid prediction of the dam break flow by using the trained prediction model. According to the method, high-fidelity flow field detail reconstruction is realized in flow field prediction based on an improved ResUNet encoder-decoder structure.

The invention discloses an image segmentation network construction method fusing visual Mama and multi-scale features, and the method comprises the steps: designing an MFVM-Unet based on a U-Net network structure, and enabling the MFVM-Unet to serve as an image segmentation network; wherein the MFVM-Unet comprises an encoder, a decoder and a multi-scale feature information guiding module; the encoder is used for gradually extracting multi-level features of an input image; the decoder is used for recovering the spatial resolution of the feature map; and the multi-scale feature information guiding module is used for integrating multi-level features. According to the method, the extraction capability of local and global features can be enhanced, and the segmentation precision of a complex-form region is remarkably improved.

The invention belongs to the technical field of artificial intelligence and data science, and discloses a causal discovery method and related device based on a large model and structural constraint, comprising the following steps: according to meta-information of each causal variable, adopting a large language model and combining multiple rounds of dialogues, multi-role play and auxiliary cue word modes, generating an initial potential causal graph of each causal variable; repeating the following steps until a preset condition is met, and outputting a candidate causal graph with the highest value score: generating the candidate causal graph by adopting a policy network based on an encoder-decoder architecture, and carrying out iterative optimization on the candidate causal graph by adopting a reinforcement learning algorithm based on an act-critic framework; and calculating a value score of the candidate causal graph after iterative optimization according to the value network, and updating parameters of the policy network based on the encoder-decoder architecture according to the value score. The efficiency and accuracy of causal relationship recognition are remarkably improved, and a causal relationship graph with high confidence can be generated in a complex data environment.

The invention discloses a mandibular neural tube CBCT panoramic image segmentation network training method based on a Mama architecture, and relates to an image processing technology in the field of computer vision. According to the segmentation network training method, an encoder and decoder architecture is adopted, tubular structure features are extracted by enlarging a receptive field and learning deformation based on feature extraction operation of a tubular convolution kernel, and then a three-vision feature fusion strategy is carried out to further guide a model to supplement attention to key features from different angles; the encoder part models global and multi-scale features through a Mama architecture, and keeps efficient in training and reasoning processes, and finally enables the network to have a more complete segmentation effect on a CBCT image of a tubular structure.

The invention discloses a medical image segmentation method based on an edge-guided attention mechanism, and the method carries out the segmentation of a medical image through a multi-scale edge-guided attention image segmentation model. The multi-scale edge-guided attention image segmentation model comprises an encoder, a decoder, an edge-guided attention module and a multi-scale feature map fusion module; by introducing an edge-guided attention mechanism into a multi-scale edge-guided attention image segmentation model, compared with an existing method, the method has the advantages that more edge details can be reserved when an image with a small target and a blurred boundary is segmented; the effectiveness of accurate recognition and boundary form segmentation of colorectal cancer osteoblast metastasis tumor by a multi-scale edge guide attention image segmentation model is improved. Meanwhile, the problems that in the colorectal cancer osteogenesis metastasis tumor segmentation process, correlation between the features is not fully considered and important features are not effectively screened in an existing method are solved, and therefore the accuracy of image segmentation is improved.

The invention discloses a radar echo extrapolation method and system based on frequency domain enhancement, and the method mainly comprises the following steps: obtaining and preprocessing a historical radar echo grayscale image sequence, generating a sequence sample through a sliding window, and dividing the sequence sample into a training set, a verification set and a test set; the method comprises the following steps: constructing a frequency domain enhanced U-Net network comprising an encoder-decoder structure, introducing a multi-scale deep convolution structure into an encoder and a decoder, and enhancing frequency domain features by using a frequency domain dynamic attention mechanism in jump connection; inputting the training set into the model for training by adopting a composite loss function comprising intensity weighted loss, frequency domain consistency loss and structural similarity loss; and inputting the test set into the trained model, and outputting a radar echo prediction result at a future moment. The method can be effectively applied to the fields of short-term and temporary weather forecast, severe convection monitoring and the like, and provides more accurate and reliable radar echo prediction support for meteorological disaster early warning.

The invention relates to the technical field of virtual power plant fault early warning, in particular to a virtual power plant fault early warning method based on a hierarchical interactive causal graph Transform, which comprises the following steps: establishing a three-level hierarchical interactive causal graph; according to a multi-resolution causal distillation mechanism, global causal association modeling is carried out on an encoder-decoder architecture of the Transform; a multi-level contrast causal attention mechanism is introduced into a Transform model for automatic quantitative analysis; performing false correlation elimination on the inter-layer complex causal relationship according to a high-order Markov random field framework; constructing an interlayer information transfer mechanism model based on the conditional variational graph diffusion model and the actual interlayer complex causal relationship; optimizing the interlayer information transmission mechanism model; and identifying a potential cascade fault path of the virtual power plant according to the optimized interlayer information transfer mechanism model. According to the method, the potential cascade fault path can be accurately identified, and the early warning accuracy of the virtual power plant in a complex fault scene is remarkably improved.