467 results about "Perceptron" patented technology

The invention provides an ocean wind field prediction method based on a neural network, and belongs to the technical field of ocean wind field prediction.The method comprises the steps that sparse ocean observation data are collected, a spatial covariance matrix is established, the spatial covariance matrix is converted into a graph structure, and then multi-hop neighborhood feature aggregation is conducted through a graph convolutional network; a tensor decomposition algorithm is combined for modeling high-order feature interaction to generate a gridding wind field, a bidirectional long-short-term memory network encoder is used for extracting space-time invariant features, a multi-layer perceptron predictor is used for directly mapping a future multi-step wind field, and a course learning strategy and a Shenchang differential equation boundary layer are matched for correction. The technical problem that sparse ocean observation data are difficult to accurately reconstruct into a high-resolution gridding wind field is solved.

The invention belongs to the technical field of machine learning and pre-stress, and discloses a pre-stress release loss value prediction method and system based on machine learning, and the method comprises the steps: carrying out the multi-working-condition modeling and simulation of a pre-stress beam through finite element numerical software, and extracting the working parameters and design parameters of the pre-stress beam, carrying out data preprocessing, distribution check and feature importance analysis to obtain an initial data set; dividing the initial data set into an initial training data set and an initial test data set, and processing the initial training data set and the initial test data set to obtain a processed training data set and a processed test data set; constructing a full-connection multi-layer perceptron neural network model, defining training, verification and monitoring functions, training the full-connection multi-layer perceptron neural network model by using the processed training data set, and testing the trained model by using the processed test data set to obtain a prediction model; real parameters of the prestressed beam are obtained, the prediction model is used for predicting the prestress release loss value, and a prediction result is obtained.

The invention belongs to the technical field of three-dimensional scene reconstruction, and particularly discloses a multi-view three-dimensional Gaussian densification method and system for adaptive density control, and the method comprises the following steps: collecting a multi-view original image, and carrying out the preprocessing of the multi-view original image; complexity features are extracted, a pixel-level complexity heat map is generated, and a globally unified three-dimensional complexity field is constructed; performing back projection on the reconstruction residual error, high-frequency inconsistency and depth / geometric consistency cost of each view angle, generating three-dimensional error popularity, determining a candidate newly-added set and a candidate pruned set, generating a weak label to train a lightweight multilayer perceptron classifier, outputting a ternary probability corresponding to newly-added / pruned / maintained, and obtaining a new / pruned / maintained three-dimensional perceptron classifier; and performing Gaussian densification operation on the newly added region. By adopting the technical scheme, fine point adding is carried out on the complex area, effective pruning is carried out on the simple area, and meanwhile, the synthesis quality, the global consistency and the calculation efficiency of the new view angle are improved.

The invention provides an intelligent inspection risk assessment method and system based on multi-sensor fusion, and the method comprises the steps: obtaining original multi-source data of a transformer substation, the original multi-source data comprises a binocular vision image, infrared thermal image data and environment sensor data, and carrying out the time-space calibration and preprocessing of the original multi-source data, obtaining a multi-source sensor data stream; carrying out feature extraction on the multi-source sensor data stream to obtain a multi-modal feature set, generating a refined semantic mask based on the multi-modal feature set, constructing an initial scene relation graph, calculating a risk level based on a multi-layer perceptron classifier, and generating a risk level evaluation result and a risk distribution graph; outputting a safety distance violation warning and a risk area identifier; and generating comprehensive risk early warning information based on the risk level assessment result, the risk distribution diagram, the safety distance violation warning and the risk area identifier.

The invention discloses a traffic flow prediction method and system based on multi-scale dynamic decomposition and a space-time Transform. According to the method, firstly, an original traffic flow sequence is decomposed into trend components and seasonal components; then, modeling the trend components by adopting a multi-layer perceptron to capture global changes; and meanwhile, a space-time Transform is used for modeling seasonal components, and the architecture effectively extracts dynamic space-time dependence characteristics by integrating space-time adaptive embedding and an adaptive Switch GLU gating mechanism. And finally, fusing trend and seasonal feature representation to generate a prediction result. According to the method, noise is effectively separated through decomposition, linear enhancement space-time self-adaptive embedding, a self-adaptive Switch GLU gating mechanism and a unified space-time self-attention Transform architecture are integrated, the modeling capacity for complex space-time dependence is enhanced, prediction precision and robustness are remarkably improved, and the method can be widely applied to the field of intelligent traffic management and control.

The invention discloses a command and control system resource trend prediction method based on fusion of long and short time sequence characteristics. The method comprises the following steps: acquiring a public power load or similar time sequence monitoring data set, and preprocessing the data in the data set; a deep learning network model based on a TCN-Transformer hybrid model is constructed, a TCN model and a Transformer model are adopted for parallel computing to achieve feature extraction, the TCN model extracts short-term information, the Transformer model extracts long-term features, then fusion features are obtained through a cross attention mechanism and multi-layer perceptron (MLP) weighting, and finally prediction output is generated through full connection layer mapping. Taking data in the training set as input, training the constructed TCN-Transform hybrid model, and continuously optimizing the model until convergence meets a set requirement; and performing prediction by using the trained network model. According to the method, the TCN-Transform hybrid model is constructed, so that local fine-grained features are reserved, the global time trend is effectively captured, and the accuracy of command decision making is improved.

The invention discloses a TR assembly gold wire bonding process parameter prediction method based on a multilayer perceptron neural network, and belongs to the technical field of microwave device intelligent manufacturing. According to the method, an intelligent mapping model of gold wire bonding geometric parameters and radio frequency performance is constructed by fusing a multi-layer perceptron neural network and parameterized electromagnetic simulation. The method specifically comprises the following steps: generating 45 groups of samples in a process parameter space by adopting Latin hypercube sampling; obtaining an S parameter data set through batch processing electromagnetic simulation; box-Cox conversion and normalization preprocessing are carried out on the data; the method comprises the following steps: constructing an MLP neural network model of a 3-32-16-2 structure, and determining hyper-parameters by using Bayesian optimization; and after training is completed, rapid reverse mapping from target performance to process parameters is realized. According to the method, the number of traditional tests is reduced from more than 200 to 45, the predicted root-mean-square error of S21 is smaller than or equal to 0.12 dB, the determination coefficient is larger than or equal to 0.96, and the parameter backstepping time lt is obtained; according to the method, full-process automation from simulation, training, optimization to production and issuing is realized, and the development efficiency of the TR component is remarkably improved.

The invention relates to a wind power cluster short-term power prediction method and device of a space-time diagram neural network fused with physical information and computer equipment, and the method comprises the steps: obtaining related information data of each wind power plant in a wind power cluster, and carrying out the preprocessing; forming a physical prior data set through an engineering analysis model fusing the wake flow analysis model and the blocking effect model; taking each wind power plant as a node of the graph, constructing graph structure data for predicting the power of the wind power plant, and forming a dynamic adjacent matrix; constructing a space-time diagram neural network WB-STGNN model architecture comprising a diagram convolutional neural network module, a gating time convolutional network and a multi-layer perceptron; the method comprises the following steps: pre-training by using a physical prior data set, and then performing formal training based on historical power data and a dynamic adjacency matrix to obtain a space-time diagram neural network WB-STGNN model; inputting the wind speed of the prediction day, and predicting the active power of the whole wind power cluster in 24 hours of the prediction day. By adopting the method, the precision and efficiency of wind power cluster power prediction can be effectively improved.

The invention belongs to the computer vision technology, and particularly relates to a small sample multi-modal prototype learning-oriented remote sensing scene classification method, which comprises the following steps of: acquiring RGB (Red, Green and Blue) images with category labels and text prompts of the RGB images as a support set; establishing a text prototype, an RGB prototype and a hyperspectral prototype of each category according to the support set; and extracting to-be-classified query set image features by using a pre-trained CLIP image encoder, calculating cosine similarities between the query set image features and the text prototype, the RGB prototype and the hyperspectral prototype of each category of the support set, taking the cosine similarities as input of a multi-layer perceptron, and obtaining the category of the to-be-classified RGB image through classification of the multi-layer perceptron. High-precision and high-robustness remote sensing scene classification is realized under the small sample condition, only prototype and similarity calculation is needed in the reasoning stage, and deployment and expansion are easy.

The invention provides a reservoir pressure salinity scheduling method based on a physical information neural network (PINN). The reservoir pressure salinity scheduling method comprises the following steps: Step 1, salinity prediction based on a physical information neural network (PINN) model; the method comprises the following steps: in a PINN model framework, embedding a physical law of salinity conservation into a multi-layer perceptron (MLP) artificial neural network for training so as to carry out salinity prediction; step 2, establishing an estuary water-salt model based on a three-dimensional ocean numerical FVCOM model; step 3, carrying out upstream reservoir salty water pressing and light water supplementing emergency scheduling based on salinity forecast; comprising the following steps: Step3.1, establishing a multi-objective function; and Step3.2, solving a scheduling model. According to the method, the salinity momentum conservation constraint can be considered, the salinity prediction accuracy under the influence of different upstream flows and downstream tidal ranges can be improved, and the emergency scheduling of the reservoir for the salt tide invasion of the estuary drinking water source can be carried out in combination with the scheduling model, so that the water supply safety is ensured.

The invention belongs to the technical field of artificial intelligence, and discloses a visual large language model illusion relieving method and a related device, and the method comprises the steps: obtaining the visual perception guide direction of each multi-layer perceptron layer of a visual large language model; and obtaining a to-be-processed visual image, reasoning the to-be-processed visual image by adopting the visual large language model, and intervening potential features output by each multi-layer perceptron layer of the visual large language model according to the visual perception guide direction in the reasoning process to obtain hallucination relief reasoning output of the to-be-processed visual image. Potential features output by multiple perceptron layers of a visual large language model are intervened according to the visual perception guiding direction in the reasoning process, the illusion problem is solved in the mode that analysis is conducted firstly and then relieving is conducted, effective intervention guidance can be provided for potential feature intervention, and therefore the intervention effect is effectively improved; and finally, object illusion in the generated content of the visual large language model is effectively reduced, and the accuracy and reliability of output of the visual large language model are improved.

The invention discloses a seawater temperature, salinity and sound velocity inversion method based on a Brillouin laser radar, and belongs to the technical field of marine environment monitoring. According to the method, the temperature, salinity and sound velocity of seawater are inverted by measuring the Brillouin frequency shift and the Brillouin line width and combining the MLP neural network; the method specifically comprises the steps of collecting seawater parameter data of different sea areas, different seasons and different depths and corresponding frequency shifts and line widths; carrying out data preprocessing and standardization on the collected data; and establishing an inversion model by using an MLP neural network, and outputting temperature, salinity and sound velocity. The multi-layer perceptron MLP neural network processes frequency shift and line width data measured by the Brillouin laser radar, high-precision and real-time inversion of seawater temperature, salinity and sound velocity is achieved, the parameter range is remarkably expanded, the inversion precision is improved, and efficient and reliable marine environment monitoring is provided.

The invention discloses a three-dimensional medical image segmentation model training method, segmentation method and system, and the method comprises the steps: constructing the architecture of a segmentation model, which comprises a self-adaptive hybrid encoder, a decoder and a classification head which are connected in sequence; a multi-layer perceptron hybrid module is introduced into the adaptive hybrid encoder to realize global feature modeling; other layers of feature maps except the last layer of feature map of the self-adaptive hybrid encoder are further processed by a residual block and a channel attention module, and each layer of the decoder performs up-sampling on the input feature map and then fuses the input feature map with the feature map of the corresponding layer of the self-adaptive hybrid encoder; obtaining the labeled three-dimensional medical image to construct a sample data set; and training the segmentation model based on the sample data set to obtain a final three-dimensional medical image segmentation model. According to the method, global context information and local detail features can be captured at the same time, and the efficiency and segmentation precision of the model are greatly improved.

The invention discloses a multi-article binding recommendation method based on graph contrast learning, and relates to the technical field of e-commerce application. According to the method, through heterogeneous binding relation modeling, a user-article-binding ternary interaction graph is constructed, direct preference of a user to an article and indirect association of binding driving are jointly learned, and the problem of relation splitting caused by multi-graph splitting in AMCBR and MGCRL is solved; according to the method, light-weight projection head optimization is carried out, double-layer linear projection is adopted to replace a traditional multi-layer perceptron (MLP), the over-smoothing risk is reduced while the contrast characterization distinction degree is enhanced, and the redundancy problem caused by fusion or knowledge modules in MultiCBR and KEMB-Rec is solved; the method introduces a multi-objective loss function, combines main task recommendation loss, cross-view comparison loss and binding comparison loss, and collaboratively optimizes local and global embedding representation, is more comprehensive than EBRec single view loss, and effectively alleviates sparsity and over-smoothing problems.

The invention relates to the technical field of artificial intelligence, can be applied to business system platforms of financial science and technology, medical health and the like, and discloses a visual feature and language feature deep fusion method, device, equipment and medium. Performing dimension mapping on the first visual feature sequence by utilizing a preset multi-layer perceptron and the text feature dimension of the to-be-fused text to obtain a second visual feature sequence, performing word segmentation and embedding processing on the to-be-fused text, and splicing the to-be-fused text with the second visual feature sequence on the sequence dimension to obtain a to-be-fused text; and carrying out self-attention analysis and fusion of a text modal path and a visual modal path on the spliced fusion feature sequence, splitting the obtained fusion attention output into a text feature part and a visual feature part, and carrying out feedforward processing and feature fusion to obtain a target visual language fusion feature. According to the invention, the cross-modal fusion efficiency and stability of vision and language are improved.

The invention relates to a method and device for predicting online open course learner satisfaction and electronic equipment, and the method comprises the steps: predicting the online open course satisfaction of students through an MLP and RBF neural network model by using a virtual learning environment of a large-scale online teaching and learning platform and combining learning behavior data in a learning management system (LMS); the model comprises a data acquisition and processing module, a multilayer perceptron (MLP) module, a radial basis function (RBF) neural network module, a classification tree module and a control block, the data acquisition and processing module is used for generating a training and testing data set, and the MLP and RBF neural network model predicts the satisfaction degree of a learner. The MLP model carries out feature extraction through a multi-layer perceptron structure and different activation functions, the RBF model measures the distance between input data and a center by using a radial basis function to realize feature extraction, the classification tree is used for judging a prediction model to which a data point belongs, the control block integrates features from the MLP and RBF neural network models, and the RBF model is used for determining a prediction model to which the data point belongs. Experimental results show that the prediction accuracy of low-satisfaction-degree learners and high-satisfaction-degree learners can be improved at the same time through the combination scheme of the MLP and the RBF, the method can be applied to learner satisfaction degree prediction of various online open courses, an educational institution is helped to know the satisfaction degree condition of students in time, course design and teaching strategies are optimized, and the teaching efficiency is improved. And important support is provided for teaching reform and optimization in the field of online education.

The invention discloses a multi-modal field adaptive false information detection method and device, and the method comprises the following steps: S1, combining a text part and an image part of a news sample based on a large visual language model and a large language model, and generating comment information; s2, extracting features of news samples from three perspectives of images, texts and multiple modes; s3, extracting comment features of the comment information, and setting a hierarchical comment guide fusion mechanism to fuse the comment features with text features and visual features respectively; s4, constructing a hybrid expert network to extract and fuse specific domain knowledge and other cross-domain knowledge of the news sample; carrying out single-view-angle prediction based on the features of each view angle and carrying out adaptive instance normalization; and performing weighted splicing on the features of different visual angles to obtain an aggregation feature, and inputting the aggregation feature into a multi-layer perceptron to obtain a classification detection result. The method has the technical effect of accurate cross-modal semantics.

The invention relates to the technical field of oyster meat yield prediction, and particularly discloses an oyster meat yield prediction method based on multi-modal fusion, and the method comprises the steps: collecting an original oyster image, and obtaining apparent data and shape parameters through a segmentation network; respectively carrying out feature extraction through a self-attention mechanism and a variational auto-encoder to obtain feature vectors; splicing and fusing the feature vectors based on a Concat method, and constructing a fused feature vector; constructing a multi-layer perceptron regression prediction model by taking the fusion feature vector as input and taking the oyster meat percentage as target output; and carrying out validity verification on the trained prediction model by using the test set. According to the method, feature extraction and splicing fusion are performed through the apparent images and the shape parameters of the oysters, the prediction model suitable for the oyster meat percentage is constructed, interaction rules of different dimensions among the features can be better captured, the model prediction precision is remarkably improved, and the problems of low manual sorting efficiency and large measurement errors in the prior art are avoided.

The invention discloses a data knowledge dual-driven unsupervised ship navigation mode identification method, and belongs to the technical field of ship traffic management. Comprising the following steps: acquiring AIS historical data sets of all ships; performing trajectory clustering based on a DBSCAN clustering algorithm of particle swarm optimization to obtain similar ship matching pairs; constructing a ship macroscopic navigation mode encoder based on a frequency domain multilayer sensor and a comparative learning method; constructing a ship microcosmic navigation mode decoder based on the hypergraph neural network; and acquiring real-time AIS data of a ship, and inputting the real-time AIS data into the trained ship macroscopic navigation mode encoder and the trained ship microcosmic navigation mode decoder to obtain intermediate features of a ship encountering scene. According to the method, the navigation mode of the ship in a habitual airway and the collision avoidance behavior mode in a multi-ship encounter scene are effectively extracted, and the limitation that the ship navigation mode recognition precision is insufficient in a single scale in the past is improved.

The invention provides a temperature compensation error elimination method in a full-automatic transformer transformation ratio test, and belongs to the technical field of transformers. A heat conduction coefficient matrix, a convective heat transfer coefficient matrix and a radiation form factor matrix are utilized to construct a temperature field dynamic evolution equation set, and a neural network compensation model combining a multi-layer perceptron architecture and an attention mechanism is adopted. A hierarchical fusion weight is dynamically adjusted based on a thermal gain stability matrix eigenvalue, a temperature gradient vector modulus length and a thermal capacity matrix condition number through a gating weight function, compensation parameters are adjusted in advance by using a temperature field prediction algorithm, and continuous improvement of measurement precision is realized through a temperature compensation effect evaluation mechanism and a dynamic parameter optimization strategy. The technical problems that the measurement precision is reduced and accurate temperature compensation cannot be realized due to temperature change in the transformation ratio test process of the transformer are solved.

The invention discloses a dementia identification method based on brain computer network space cross attention fusion. The dementia identification method comprises the steps that resting-state electroencephalogram signals are acquired, preprocessing and brain network construction are carried out, the frequency band power ratio is calculated, and a training set and a test set are divided; the brain network fuses the spatial information, and spatial features are obtained through an isotropic graph neural network and a local feature enhancement module; the frequency band power ratio is coded by a multi-layer perceptron to obtain frequency spectrum statistical characteristics; a bidirectional cross attention module is input, and complementary fusion features are extracted; and performing mixed pooling and splicing on the complementary fusion features, then sending the fused features to a Chebyshev Kolmogorov-Arnold network classifier, and outputting an Alzheimer's disease / frontotemporal dementia / health control (AD / FTD / HC) classification result and a cognitive scale evaluation score (MMSE). According to the method, the multi-feature complementary information is effectively integrated through joint modeling of the spatial diagram features and the global spectrum features, and the accuracy, stability and generalization ability of AD and FTD classification in a complex brain network are improved.

The invention provides a multi-scale implicit-explicit mixed medical image registration method and system. The method comprises the following steps: acquiring CT images of a moving image and a fixed image; processing the CT image by using the trained registration network model, performing down-sampling on a three-dimensional coordinate grid to different set proportions of an original resolution, and then predicting deformation of different scales by using a multi-layer perceptron network model to generate a first deformation field and a second deformation field; down-sampling the fixed image and the moving image to a third set proportion of the original resolution, predicting third scale deformation by using the CNN network model, and generating a third deformation field; and fusing the three deformation fields to obtain a final deformation field, and generating a final deformation image through a spatial transformation network. The method is simple in registration process, and has good practical application efficiency and expandability.

The invention discloses a self-adaptive multi-phase parallel power supply management system, and relates to the technical field of intelligent regulation and control of a power system, and the system comprises a feature extraction module which is used for extracting periodicity, tendency and randomness features from historical power utilization data and real-time sensor data, determining a load sudden increase probability through a time sequence decomposition algorithm, and determining a load sudden increase probability; obtaining a load sudden increase probability value; the load prediction module is used for processing real-time load data through a multi-layer perceptron structure if the load sudden increase probability value is higher than a preset threshold value, predicting a load sudden increase peak value and duration in an hour-level or minute-level time window, and obtaining a load sudden increase prediction result; according to the self-adaptive multi-phase parallel power supply management system, multi-algorithm collaboration and real-time dynamic optimization are adopted, the operation stability and power supply continuity of a power grid are remarkably improved, and overload and voltage fluctuation risks are reduced.

The invention relates to the technical field of medical health risk assessment, and particularly discloses a coronary heart disease risk assessment method fusing a tongue diagnosis image and structured data, and the method comprises the steps: collecting the tongue image of a patient with coronary heart disease and the structured data such as age and gender; and extracting tongue image features by using a convolutional neural network, encoding the structured data through a multi-layer perceptron, and fusing the encoded structured data with the structured data. And a CNNGPT2 deep learning model and a CNNTabular deep learning model are constructed. After training, a test set prediction result is extracted to construct a metadata set, XGBoost is used as a meta classifier for training and prediction, a coronary heart disease early prediction model is constructed, and coronary heart disease risk assessment is performed based on the early prediction model. According to the method, the tongue diagnosis image and the structured data are fused, the limitation of a single data type is made up, and the comprehensiveness of evaluation is improved. According to the method, two deep learning models of CNNGPT2 and CNNTabular are constructed, association information between data is fully mined, and the processing capability of complex data is improved.

The invention relates to the field of automatic driving system test and evaluation, in particular to an automatic driving automobile intelligence degree evaluation method based on subjective and objective mapping of a large language model, and can overcome the defects of a traditional evaluation method in the aspects of objectivity, efficiency and nonlinear fitting ability. The method comprises the following steps: converting automatic driving interaction data into natural language description; constructing an automatic driving field knowledge base, and performing knowledge enhancement on the large language model to improve the understanding of the large language model on professional rules; performing quantitative evaluation on interaction data in multiple dimensions (safety, comfort, efficiency, social interactivity and influence on a traffic system) by using the enhanced large language model; a nonlinear mapping model from interaction data to intelligent scoring is established by adopting a multi-layer perceptron, and efficient and objective comprehensive evaluation is realized. The method can effectively reduce the dependence on artificial experts, improves the evaluation consistency and efficiency, and is suitable for the intelligent level comprehensive evaluation of different levels of automatic driving systems.

The invention belongs to the technical field of fetal brain age prediction, and relates to an earthworm cerebellar fetal brain age prediction method and system based on multi-modal feature fusion, an MST-Mamba segmentation network is adopted, and local-global aggregators are embedded in each level of an encoder, so that the cooperation of local detail capture and global semantic modeling is realized; meanwhile, a dynamic channel fusion device is deployed at the jump connection part of the encoder and the decoder, so that the problems of fuzzy boundary, missed division, wrong division and the like are avoided; through three parallel branches of a multi-granularity form-texture collaborative perception architecture, two types of explicit features of macroscopic geometry and topological form and implicit features of microscopic texture are synchronously extracted, and comprehensive characterization of the development features of the earthworm cerebellar part is realized; the explicit features are subjected to standardized calibration and then spliced and fused with the implicit features in the channel dimension, the problems that multi-modal feature fusion is insufficient and calibration lacks are solved, finally prediction is conducted through a multi-layer perceptron regression head, and the accuracy and stability of the brain age prediction result are guaranteed from the source.

The invention relates to an automatic instrument parameter self-tuning method and device based on deep learning. According to the method, real-time operation data of an instrument is collected through a multi-source sensor, time sequence alignment is carried out to form a multi-dimensional data tensor, and then local time sequence features and a long-term dependency relationship are respectively captured by using a convolutional layer and a bidirectional long-short-term memory network in a depth feature extraction network; the two types of features are fused through an attention mechanism to form a depth feature vector, on this basis, the vector is mapped into a working condition membership degree vector by adopting a differential fuzzy clustering method, and finally, the working condition membership degree is non-linearly mapped into a PID parameter adjustment amount through a multi-layer perceptron network, and the PID parameter adjustment amount is superposed to a basic parameter to realize parameter self-tuning. Therefore, under the condition that manual intervention is not needed, an automatic instrument can automatically adapt to complex and changeable operation conditions, the control precision and the system stability are remarkably improved, and the problem of adaptability of traditional PID control in a time-varying nonlinear system is effectively solved.

The invention provides a multi-modal sentiment analysis model and method, electronic equipment and a medium, and the model comprises a feature enhancement module which is used for extracting original multi-modal data features through an exclusive tool, constructing a graph structure, and enhancing the graph structure through a graph convolutional network to obtain multi-modal enhanced features; the modal multi-stage balance module is used for processing enhanced features by using different multi-stage network structures and outputting multi-modal consistency representation; the modal noise reduction decoupling and specificity recombination module is used for obtaining low-noise representation through a modal noise reduction decomposer based on the global information and recombining the low-noise representation in a modal bank to generate low-noise multi-modal specificity representation; and the hierarchical fusion prediction module is used for fusing consistency and specificity representation according to single-peak, double-peak and three-peak modes, and outputting an emotion prediction result through a multi-layer perceptron.

The invention discloses a transport capacity resource integrated intelligent scheduling method and system oriented to multiple service scenes, and relates to the field of intelligent scheduling, and the method comprises the steps: constructing a hierarchical collaborative decision-making architecture comprising a macroscopic strategic layer agent and a microscopic tactical layer optimizer; and inputting the structured feature vector into a hierarchical collaborative decision-making architecture, dynamically distributing proper solution algorithms and parameters for a microscopic tactical layer optimizer according to a real-time scheduling situation by utilizing an online element learning optimizer, and outputting a pre-scheduling scheme. According to the invention, through integrated integration of multi-service scene data, comprehensive utilization of static basic information, real-time transport capacity data and prediction environment data is realized, and the data support capability of scheduling decision is improved. The feature interaction network of the multi-layer perceptron structure can accurately extract core features and provide effective input for scheduling decisions. The hierarchical collaborative decision-making architecture is combined with an online element learning optimizer, so that a solution algorithm and parameters can be dynamically matched, and a better pre-scheduling scheme can be output.

The invention provides an urban environment performance prediction method and system based on multi-modal fusion and attention enhancement, and the method comprises the steps: collecting image data and numerical data, and generating a plurality of urban environment performance distribution truth value maps; preprocessing the multi-modal data to obtain an image map and a numerical value feature vector, and pairing the image map and the numerical value feature vector with the true value map to form a multi-modal data set; a multi-target model of a conditional generative adversarial network based on attention enhancement is constructed, a generator of the multi-target model comprises a two-way encoder, spatial features can be extracted based on an image map, and physical features can be extracted based on a numerical feature vector; an image coding path adopts a U-Net down-sampling structure containing a convolution block attention module, and a numerical value coding path adopts a multi-layer perceptron; and the multi-head output layer outputs a plurality of predicted urban environment performance distribution diagrams. After the model is trained, target area data are input, and three types of prediction distribution diagrams are output. According to the invention, the urban environment performance prediction effect is improved.