586 results about "Autoencoder" patented technology

This invention relates to the field of digital management technology for engineering projects, and discloses an intelligent engineering performance evaluation management system and method based on job-specific dimensions. The system includes: acquiring engineering task data and time-series data of external environmental conditions; generating a task-dependent directed acyclic graph and labeling directed edges according to time segments; calculating the conditional control quality transmission influence coefficient matrix for each time segment based on partial correlation analysis; tracing indirect contribution paths along the directed acyclic graph and calculating condition-perceived indirect contribution values; fusing direct and indirect contribution values ​​to generate a full-link contribution degree; inputting the full-link contribution degree sequence and the external environmental condition time series into a time-series conditional variational autoencoder to extract a potential capability space vector; generating a counterfactual expected total performance; and fusing capability quantile values ​​and effort indicators to generate a comprehensive performance evaluation score.

This invention provides a method, device, and medium for sensing abnormal features of photovoltaic arrays, belonging to the field of power system technology. The method includes: acquiring photovoltaic array operating data; performing distribution correction on the photovoltaic array operating data to obtain corrected photovoltaic array data; inputting the corrected photovoltaic array data into a pre-constructed abnormal feature sensing model, and outputting the photovoltaic array abnormal feature sensing result. The construction of the abnormal feature sensing model includes: embedding physical quantity constraints for correcting weakly correlated physical quantities into the latent space of the original autoencoder to obtain a PCAE structure; and sequentially adding a CNN structure and an LSTM structure after the PCAE structure to obtain the constructed abnormal feature sensing model. This invention ensures that the encoded features conform to the working principle of the photovoltaic array, improves the comprehensiveness of anomaly identification and the accuracy of data state estimation, and solves the problem of parameter distribution offset.

The application belongs to the technical field of power grid transient stability evaluation, and specifically discloses a boundary sample enhancement method and system for power grid transient stability evaluation, which comprises the following steps: training a transient stability evaluation model by using an initial training sample set, predicting the initial training sample by using the trained transient stability evaluation model, and screening boundary samples based on the obtained prediction probability; calculating the local density of all non-boundary stable samples, and performing undersampling on the non-boundary samples; training a mask autoencoder generative adversarial network by using the screened boundary samples, generating new boundary samples without labels, obtaining sample labels by using a time domain simulation technology, and adding the new samples that pass the test to the training sample set after undersampling to realize boundary sample enhancement. The application can effectively identify the samples near the classification decision boundary of the transient stability evaluation model, and provides reliable training data basis for boundary sample generation and enhancement.

An operator learning model generator is provided for training a smooth operator learning model for predicting airflow dynamics in a room used by a controller connected to a heating, ventilation and air conditioning (HVAC) system. The operator learning model generator includes an interface circuit configured to receive a training dataset via a network connected to a simulation computer, wherein the training dataset includes solution trajectories of airflow in the room for various times series of control actions given to the HVAC system, a memory configured to store the smooth operator learning model comprising an auto-encoder and a neural ordinary differential equation, the training dataset, and training instructions for the smooth operator learning model, and a processor configured to train the smooth operator learning model stored in the memory, wherein the training instructions comprise a jerk regularization that enforces smoothness of the dynamics predicted by the smooth operator learning model.

This invention belongs to the field of bioinformatics and relates to a single-cell trajectory inference method based on adaptive feature selection. First, an initial gene expression matrix is ​​obtained through data preprocessing and screening for highly variable genes. Second, a two-dimensional evaluation strategy is employed to calculate the scores of highly variable genes with gene expression variability and the trajectory importance score related to differentiation trajectories. Then, a dynamic weight fusion mechanism is introduced, adaptively adjusting the fusion weights of the two scores based on performance feedback, and highlighting key genes through nonlinear enhancement. Next, an intelligent inflection point detection algorithm adaptively determines the optimal number of features. Finally, trajectory inference is performed based on a variational autoencoder model reconstructed from feature subsets, and a performance-driven feature selection closed loop is formed through multiple rounds of iterative optimization. This invention achieves high-precision, adaptive single-cell trajectory inference, solving the technical problems of single feature selection and fixed weights in traditional methods.

An operator learning model generator is provided for training a smoothed operator learning model for predicting airflow dynamics in a room used by a controller connected to a heating, ventilation, and air conditioning (HVAC) system. The operator learning model generator includes an interface circuit configured to receive a training dataset via a network connected to a simulation computer, wherein the training dataset includes un-trajectories of airflow in the room for various time series of control actions applied to the HVAC system; a memory configured to store the smoothed operator learning model including an autoencoder and a neural ordinary differential equation, the training dataset, and training instructions for the smoothed operator learning model; and a processor configured to train the smoothed operator learning model stored in the memory, wherein the training instructions include jerk regularization that constrains smoothness of dynamics predicted by the smoothed operator learning model.

The application provides a multi-view clustering method and system based on a graph attention automatic encoder, relates to the technical field of multi-view clustering, and specifically includes the following steps: selecting a view with the largest information quantity from different views of the same group of nodes; learning a graph structure and node content by using a trained graph attention encoder based on the view with the largest information quantity and node content information, so as to obtain a node feature representation; performing specific constraint on the node feature representation by using an l1,2-norm penalty, so as to obtain a constrained node feature representation; inputting the constrained node feature representation into a self-optimizing clustering module to perform clustering, so as to obtain a final clustering result; and the application applies the graph attention network to multi-view graph clustering, simultaneously reconstructs the graph structure and the node content, and makes the latent representation well preserve the graph structure and the content information of the nodes, so that the application is more suitable for cluster tasks.

The embodiment of the application relates to the technical field of dance generation, and provides a dance generation model training method and a dance generation method, a deep-coupled neural network architecture is constructed, a dynamics constraint module based on a space-time graph neural network is innovatively introduced in a latent space of a variational autoencoder, specific dynamics bias is explicitly predicted and applied to an action intention according to a specified clothing type, dance action sequences conforming to physical laws and having specific clothing dynamics are forced to be generated by the network output, the generated dance action sequences can not only match music rhythm, but also can truly present physical dynamics of the specified clothing, and the authenticity and expressiveness of dance generation under complex clothing are significantly improved.

This application discloses a method, apparatus, equipment, and storage medium for locating remaining oil. The method includes: acquiring segmented displacement state data of well groups and boundary transition trajectory data of the development stage; obtaining a continuous potential displacement state sequence by boundary interception, trajectory insertion, and time concatenation; obtaining a continuous displacement state sequence over all time periods by using a three-layer fully connected decoding method mirroring the encoding layer of the preceding depth Koopman autoencoder network; extracting the state at six consecutive sampling times at the boundary to form a local continuous change sequence; obtaining displacement obstruction judgment parameters through two-layer fully connected compression mapping; generating two types of markers through threshold comparison, cross-screening contradictory well groups, associating spatial locations, and merging to obtain the location result. This application can accurately distinguish two types of easily confused areas, and the location result can directly support the selection of sites and potential tapping decisions for later-stage oilfield development measures.

The application discloses a work clothes identification method combining continuous learning and effectively resisting forgetting disaster, and comprises the following steps: collecting data in an initial stage, constructing a work clothes classification model, and combining data to perform self-supervised training on the work clothes classification model; the trained work clothes classification model is divided into a deep model and a shallow model, an image self-encoder is constructed between the deep model and the shallow model, and the image self-encoder is trained, the intermediate features output by the shallow model are encoded and compressed through the trained image self-encoder, and compressed features are obtained; in a continuous learning task stage, work clothes category data are collected, the compressed features are decoded, and a new model is constructed; based on the new model, a Tt-stage work clothes classification model is constructed through a gradient boosting method, and after training, compression is performed to obtain a final model for work clothes identification; the application alleviates the forgetting degree of the work clothes classification model to old categories, and improves the plasticity of the model.

A system for dynamic latent space adaptation using spatiotemporal kernel context for multiscale rendering with hierarchical and Lorentzian autoencoders. The Spatiotemporal Kernel Estimator (SKE) analyzes media through motion field, temporal recurrence, frequency band, and scene semantics analyzers to generate adaptive kernel parameters encoding content-specific importance distributions. The system dynamically adapts latent manifold geometry by modifying metric tensor properties according to kernel context, enabling content-aware compression that allocates representational capacity based on visual significance. A multiscale cache implements kernel-adaptive retention policies prioritizing important regions. An adaptive renderer provides intelligent level-of-detail selection based on zoom level and kernel-estimated importance, optimizing processing allocation. The self-optimizing architecture continuously refines kernel context and geometric adaptation based on user interaction and performance feedback, achieving superior compression ratios and perceptual quality. Applications include bandwidth-efficient video streaming, virtual reality, scientific visualization, and cognitive video analytics requiring intelligent context-aware visual processing.

The present application belongs to the technical field of intelligent processing of electric shovels, and discloses an electric shovel intelligent excavation prompting system based on a hybrid encoder and a Transformer decoder, comprising: a data acquisition module, a sensor is arranged on the electric shovel to acquire real-time operating parameters of the electric shovel and real-time environmental information parameters of the working environment of the electric shovel; a hybrid encoder module, which dynamically constructs the relationship between nodes and edges through a graph neural network for the real-time operating parameters and the environmental information parameters, and converts the relationship into graph structure data; and removes noise information from the graph structure data through a cross-modal variational autoencoder to obtain encoded structure data; and a Transformer decoder module, which captures the correlation of the encoded structure data based on a self-attention mechanism, acquires working condition information of the electric shovel, and outputs excavation prompting information. The present application realizes intelligent processing of real-time operating parameters and environmental information parameters of the electric shovel, outputs excavation prompting information to assist the electric shovel driver in operation, and improves the efficiency and safety of electric shovel operation.

The present application belongs to the field of non-technical line loss reduction, and discloses a power stealing detection method based on stacked sparse autoencoder and deep forest, comprising: S1: extracting user power consumption data from an intelligent power supply and consumption database to construct a user power consumption feature dataset; S2: dividing the user power consumption feature dataset into a training set and a test set according to a certain proportion; S3: taking the training set as the input of a power stealing detection model to train the power stealing detection model; S4: testing the power stealing detection model by using the test set and constructing a confusion matrix according to the test result; and S5: evaluating the test result by using accuracy, recall rate and F1 score respectively. The present application uses a stacked sparse autoencoder to reduce the dimension of power consumption data, extracts effective information in high-dimensional power consumption data, and performs power stealing detection, thereby solving the problem of "curse of dimensionality" caused by high-dimensional power consumption data and improving the recognition accuracy of the power stealing detection model, so as to more effectively reduce non-technical line loss.

A non-transitory computer-readable recording medium stores therein a training program that causes a computer to execute a process including first inputting a first image capturing a target compound to an encoder of an auto-encoder including a latent space that is isometric with respect to an input space, second inputting a latent variable output by the encoder and a typical compound model corresponding to a typical case of a three-dimensional structure of the target compound to a decoder of the auto-encoder, and updating parameters of the encoder and the decoder, based on a reconfiguration error between a second image reconfigured based on an output of the encoder and the first image.

A method is provided for adapting a diffusion-based image generation model (2) comprising at least a denoising autoencoder (4). The method comprises: obtaining a style-specific noise distribution in an image space or in a latent space for a respective set of target style images; and adjusting parameters of the denoising autoencoder (4) by using the style-specific noise distribution to adjust the parameters of the denoising autoencoder (4) to obtain an adapted diffusion-based image generation model (2).

Various example embodiments relate to the expansion or compression of data in transport blocks. An apparatus may include: components for receiving training auxiliary data from another apparatus for training at least a portion of a reverse autoencoder neural network, the reverse autoencoder neural network including: an expander neural network configured to: determine an expanded representation of the transport block data such that the transport block data has a specified size; and a compressor neural network configured to: determine a compressed representation of the transport block data based on the expanded representation of the transport block data to reconstruct the transport block data, wherein the expander neural network is an encoder of the reverse autoencoder neural network, and the compressor neural network is a decoder of the reverse autoencoder neural network.

This disclosure provides a training method, a 3D model completion method, and an apparatus for a 3D model completion network, including: pre-constructing a 3D model completion network to be trained; the 3D model completion network to be trained includes a 3D variational autoencoder and a diffusion model; acquiring a 3D model to be trained for network training; inputting the 3D model to be trained into the 3D variational autoencoder, inputting the encoded latent vectors into the diffusion model, performing noise addition and denoising processing on the diffusion model, and then inputting the latent vectors into the decoder to obtain a predicted generated 3D model; based on the predicted generated 3D model, calculating the loss of the 3D variational autoencoder and the diffusion model, and training the 3D model completion network to be trained. The trained 3D model completion network is then used to complete the 3D model to be completed. This ensures the quality of the 3D model while improving production efficiency.

The application discloses a kind of high-resolution radar echo extrapolation prediction methods of fusion satellite data, specifically as follows, first, input history radar echo sequence pretreatment of previous T time, including denoising, normalization processing, data set segmentation, obtain cleaned data;Then, by deterministic modeling method (SimVP), obtain the fuzzy prediction sequence of future T length, then variational autoencoder (VAE) respectively original radar echo image and fuzzy prediction sequence are mapped to low-dimensional latent space, and two-stage diffusion modeling is carried out on this basis;For the first stage, utilize space-time converter (ST-Translator) to extract the space-time evolution characteristics of radar echo;Second stage first input corresponding time satellite data of previous T time, pretreatment is carried out, including normalization processing, feature selection, data set segmentation, obtain cleaned data, adopt multi-source fusion denoising network Fsrformer, dynamically adjust the influence of satellite data in diffusion process, to make full use of satellite information;Finally, the output result of two stages is inversely transformed to pixel space, and the high-resolution radar echo extrapolation prediction result of future T length is obtained.The application can effectively reduce the consumption of computing resources, improve the precision and detail fidelity of short-term precipitation prediction.

The application discloses a kind of real-time detection method and system of food pollutant based on intelligent sensor, comprising: through the multimodal sensor array of deployment in food processing production line key station, sensor response signal is collected;Baseline response under the condition of no pollution is predicted by establishing dynamic baseline prediction model;Signal separation network based on variational decoupling autoencoder is constructed, residual signal is mapped to low-dimensional latent space by sharing encoder and is divided into baseline error subspace and pollutant signal subspace, respectively by two independent decoders reconstructing baseline prediction error signal and pollutant candidate signal;Multi-scale time-frequency transform is carried out to pollutant candidate signal to identify pollutant type and estimate concentration;Through distributed intelligent agent cooperation, pollution traceability and hierarchical early warning are carried out;Cross-scene adaptive calibration is realized using meta-learning framework.The application realizes the real-time decoupling of pollutant signal and matrix change signal in food processing process.

We provide video coding, encoders, and decoders that further improve efficiency based on trained networks. [Solution] To handle lumer-chroma channels of different sizes, the chroma component is upsampled so that the resulting upsampled chroma component has a resolution matching one of the lumer components. The lumer and upsampled chroma components are then encoded into a bitstream. To reconstruct the picture portion, the lumer component and an intermediate chroma component matching the resolution of the lumer component are decoded from the bitstream, and then the intermediate chroma component is downsampled. The subsampled chroma format is handled by an autoencoder / autodecoder framework while preserving the lumer channel.

The application belongs to the field of biomedical imaging, and particularly relates to an EEG diffuse source imaging method based on a self-encoder and spatial difference sparse constraint, which comprises the following steps: brain modeling, calculation of a lead matrix, generation of simulated brain source signals by using a region growing method; taking the simulated brain source signals as inputs of a self-encoder, inversing the cortical brain source signals through the self-encoder, and outputting reconstructed electroencephalogram signals; calculating a loss function, and optimizing network parameters through back propagation; visualizing the reconstructed brain source signals; the loss function is composed of three parts, the first part is a mean square error between simulated and reconstructed electroencephalogram data, the second part is a mean square error and a mean absolute error between simulated and reconstructed brain source signals, and the third part is a spatial difference sparse regularization term of the brain source signals; through joint optimization of the three parts, accurate and robust cortical diffuse source imaging estimation is obtained, and technical support is provided for the field of neural disease diagnosis and brain-computer interface.

This invention relates to the field of acoustic emission signal feature enhancement for aero-engines, and discloses a method, system, medium, and device for denoising weak fault signals under strong background noise in aero-engines. The method includes: acquiring strong background noise signals under normal operating conditions of the aero-engine; performing wavelet packet decomposition and reconstruction to obtain multiple frequency band component signals; constructing a corresponding autoencoder for each frequency band component signal, and using the frequency band component signal as input to perform unsupervised training on the autoencoder; performing wavelet packet decomposition and reconstruction on the acoustic emission signal to be processed to obtain multiple frequency band component signals to be processed; inputting each frequency band component signal to be processed into the corresponding trained autoencoder for reconstruction; calculating the reconstruction error between the input and output signals of each autoencoder; determining whether each frequency band component signal contains outliers based on the reconstruction error; reconstructing the frequency band component signals containing outliers to obtain the denoised fault signal.

This application discloses a multimodal cognitive trap accumulation assembly calculation method and system, applied to embodied intelligent robots. The specific method steps include: (1) learning, comparing, and embedding text corpora in the human language domain; (2) encoding trap perception data; (3) constructing a symbolic trap library SS; (4) collecting physical interaction data through a multimodal sensor array, compressing it through a spatiotemporal graph convolutional network and a variational autoencoder to generate an embodied trap library EE; (5) training a dual-tower contrastive learning model to establish a mapping relationship between symbolic traps and embodied traps, generating a cognitive anchor library AA. By constructing the cognitive anchor library AA, trap behavior and probability of the embodied intelligent system when performing tasks are avoided or reduced.

This invention discloses a method and device for detecting the load at the charging port of an electric vehicle charging pile, relating to the field of charging safety technology. The method includes: acquiring an electrical signal from the charging port; extracting a clean load response signal using a quantum noise-resistant processor based on Majorana fermion operators; inputting the clean load response signal into a pulse neural network autoencoder to generate a neural pulse sequence through a pulse firing mechanism; performing time-frequency domain transformation on the neural pulse sequence to reconstruct a complex impedance tensor, and solving the eigenvalue sequence of the complex impedance tensor; embedding the eigenvalue sequence into a five-dimensional anti-de Sitter space boundary based on the AdS / CFT duality principle, and generating a three-dimensional holographic tomographic map by solving the Einstein field equations relating the volume to the boundary; this invention achieves topological separation of noise and load signal through a Majorana zero-mode waveguide channel, solving the problem of insufficient high-frequency noise suppression in dynamic load scenarios.

This invention discloses an improved method for generating driver fatigue EEG data using a latent diffusion model, belonging to the field of EEG signal processing. The method includes: processing the original multi-channel driver fatigue EEG signal through multi-scale wavelet denoising, adaptive thresholding, and an improved soft thresholding function; then performing inter-channel covariance alignment and whitening to eliminate redundancy to obtain a preprocessed signal; performing short-time Fourier transform and logarithmic energy normalization to obtain normalized time-frequency features; inputting the time-frequency features into an encoder to obtain the mean and variance of latent variables, and obtaining latent variables through reparameterized sampling; reconstructing the time-frequency features using a decoder combined with fatigue state labels; training a conditional variational autoencoder by minimizing reconstruction loss and KL divergence loss; adding noise through forward diffusion in the latent space; training a denoising network to remove noise based on fatigue state labels during reverse denoising; inputting the denoised latent variables into the decoder, combining them with a specified fatigue state to generate new time-frequency features, and reconstructing them into the target EEG signal. This invention can enhance the training dataset and improve the accuracy of driver fatigue monitoring.

The application provides an intelligent image processing system and method for a car based on a perception algorithm model, and the method comprises the following steps: S1. spatio-temporal reference double anchoring and dynamic intrinsic extrinsic parameter calibration of a vehicle-mounted image acquisition node; S2. multi-node image heterogeneous domain normalization and adaptive preprocessing based on a self-adaptive kernel regression non-local mean denoising algorithm; S3. hierarchical image feature extraction and semantic anchoring based on a graph neural network dynamic feature interaction network; S4. cross-node and cross-frame feature mutual checking and pseudo-feature elimination; S5. full-scene semantic completion and dynamic target trajectory prediction based on a variational autoencoder trajectory prediction model; S6. dynamic lightweight adaptation and algorithm power adaptive scheduling of the perception algorithm model; and S7. risk scene grading identification and image targeted enhancement output based on semantics and trajectories. The application provides stable, accurate and efficient vehicle-mounted image perception support for intelligent driving of a car, and improves the safety and adaptability of environmental perception of intelligent driving.

The present application relates to clustering analysis technology in data mining and high-level fusion technology in information fusion, and belongs to the field of pattern recognition and intelligent information processing. It includes: 1) setting the attributes and type labels of the target; 2) representing the target data in the form of a space-time graph; 3) designing a space-time graph autoencoder with an attention mechanism, learning node representation by aggregating adjacency matrix information, and reconstructing the space-time graph network structure by calculating the inner product of node pairs; 4) constructing a self-training graph neural network model to aggregate the reduced neighbor target information; 5) designing a graph convolutional neural network and a deep neural network double self-supervised module; step 6, setting the target behavior rule label; 6) visualizing the target behavior rule. The method can solve the problems of traditional clustering methods such as strong data dependence, high computational complexity and inaccurate measurement description, and realize efficient mining and analysis of target behavior rules.

The application discloses a safety control method and system for unmanned loader based on machine learning, and relates to the technical field of safety control.The method comprises the following steps: processing multi-source sensor data by using an anti-vibration feature extraction network to generate a safety evaluation coefficient; analyzing a safety state based on a variational autoencoder model to generate structured safety warning information; applying reinforcement learning for multi-objective optimization to output a safe driving trajectory; and finally converting the trajectory into a control instruction through a deep learning controller and adjusting the control parameters in real time based on an online learning mechanism.The application realizes intelligent safety control of the unmanned loader under complex working conditions by constructing a whole-process machine learning processing chain from data acquisition to control execution, and solves the problems of poor safety control adaptability, insufficient cooperation between control modules and vibration interference affecting the accuracy of judgment in the prior art.

Various examples of the disclosure pertain to using whole-slide images that depict healthy tissue for a training process for at least one machine-learning algorithm for digital pathology. For instance, an autoencoder neural network can be trained based on the healthy tissue.