14results about How to "Improve deployability" patented technology

This invention discloses a real-time rendering method and system for physically simulated volumetric clouds. The method includes the following steps: constructing a dynamic noise mixing model, and constructing complex cloud layers by mixing multiple noises based on the dynamic noise mixing model; constructing a multi-scattering illumination model, and using the multi-scattering illumination model to perform single-scattering calculations and multi-scattering approximations on the complex cloud layers to simulate illumination; based on the completed illumination simulation, dividing the cloud density field of the complex cloud layers into several voxel blocks, and performing adaptive light travel processing to complete the real-time rendering of volumetric clouds. This invention, without sacrificing image quality and rendering efficiency, integrates advanced cloud modeling theory with efficient GPU acceleration technology, and uses a mixed noise model to generate diverse cloud structures, accurately reproducing various typical cloud types such as high-altitude cirrus clouds, cumulus clouds, and stratus clouds.

This invention relates to a method and system for automated generation of TSN network configurations based on a reflective agent. The system comprises a network state perception and intent reflection module, a YANG model configuration parameter and structure decision module, and a YANG model intelligent generation module. The network state perception and intent reflection module is responsible for structurally perceiving network state and resource information. The YANG model configuration parameter and structure decision module integrates multi-source knowledge and contextual information, intelligently generating deployable TSN scheduling parameters and YANG configuration structures based on business needs. The YANG model intelligent generation module automatically generates the corresponding YANG configuration model based on the optimized results, performs syntax and semantic verification, and then distributes it to the target device, achieving automated deployment of network configurations. This invention significantly improves the intelligence level and deployment automation capability of TSN network configuration by introducing reflective reasoning and feedback optimization techniques.

This invention provides a method and system for predicting atrial fibrillation (AF) load based on intermittent PPG signals, belonging to the field of data processing technology. The method includes: training a first neural network using a first PPG signal with known AF results to obtain an AF probability prediction model; using the AF probability prediction model to predict the AF probability of a second PPG signal within a target time period and mapping it to an AF risk value; determining the AF probability at any given time based on the risk values ​​of two adjacent second PPG signals, integrating to obtain the total AF time, and thus obtaining the AF load; training a second neural network using historical AF loads from multiple time periods to obtain an AF load prediction model, used to predict the AF load for multiple future time periods. This invention achieves quantitative assessment and future trend prediction of AF load from discrete PPG signals to continuous AF load.

The application discloses a multi-unmanned aerial vehicle target search method based on self-attention and reinforcement learning, which converts the target search task of the unmanned aerial vehicle into a multi-agent cooperation problem, takes each unmanned aerial vehicle as an independent agent, adopts a strategy network SA-MADDPG to search for a target, simultaneously adopts a Voronoi diagram to reasonably divide a search area, and uses a target probability map TPM to help evaluate the possible distribution of the target, so that the unmanned aerial vehicle preferentially searches a high target probability area, and finally adjusts a search strategy according to real-time feedback during the execution of the search task, so as to adapt to the dynamic change of the target.

The application discloses a Mongolian-Chinese multi-expert neural machine translation method and system based on decision layer relative strategy optimization, and the method comprises the following steps: acquiring Mongolian-Chinese bilingual parallel corpus and preprocessing to obtain Mongolian-Chinese translation training data; training a basic translation model based on the Mongolian-Chinese translation training data to obtain a Mongolian-Chinese task adaptation model; performing sensitivity analysis on the Mongolian-Chinese task adaptation model to obtain hierarchical sensitivity results of each network layer; according to the hierarchical sensitivity results and the complexity of an input sentence, different network layers of the Mongolian-Chinese task adaptation model are allocated with different quantization precisions to obtain a light-weight translation model; a course type training sequence is constructed, and an optimized Mongolian-Chinese translation model is obtained based on a composite reward function; a Mongolian sentence to be translated is input into the optimized Mongolian-Chinese translation model to generate a corresponding Chinese initial translation; and a final Chinese translation is determined according to the Mongolian sentence to be translated and the Chinese initial translation.

The application discloses a double-embedding model hybrid training method, system, device and storage medium, which is applied to the technical field of recommendation systems and comprises the following steps: determining the routing parameters of each classification feature based on the historical access information of the classification features in a training data set; selecting a corresponding embedding table for the classification features input currently according to the routing parameters; wherein the embedding table comprises a first embedding table and a second embedding table; extracting the embedding vector corresponding to the classification features from the selected embedding table, and performing forward calculation and loss calculation of the model based on the embedding vector; and updating the parameters of the first embedding table, the second embedding table and a downstream model according to the result of the loss calculation. The application simulates the embedding hybrid use mode in reasoning in the training stage, enhances the collaborative ability of the two embedding tables, and improves the performance and robustness of the model in a real reasoning scene.

This invention provides a machine vision-based experimental platform for analyzing yarn twisting motion. The platform includes a support frame, an LED light, twisting devices, a support platform, and at least two cameras. The two twisting devices are fixed to the support frame and arranged opposite each other, with a yarn connected between them. The support platform is connected to the support frame and supports the cameras. The cameras are used to capture the yarn twisting motion process. The LED light is connected to the support frame and provides illumination for the yarn twisting process. This provides an experimental platform capable of acquiring three-dimensional dynamic motion data of the yarn twisting process with high precision and non-contact, thus breaking through existing research bottlenecks and laying the foundation for a deeper understanding and optimization of the twisting process.

The invention discloses a multi-unmanned aerial vehicle collaborative search scheduling method based on deep reinforcement learning. The method is oriented to a complex environment with dynamic tasks and space risk constraints. According to the method, a hierarchical collaborative decision framework is constructed, system-level task scheduling and individual-level continuous control are decoupled and modeled, a graph structure and an attention mechanism are introduced into a scheduling layer, and global optimization distribution of multiple unmanned aerial vehicles and multiple tasks is realized; in a control layer, a deep reinforcement learning strategy is adopted to generate a continuous control action based on local observation, and the coordination of scheduling and execution is improved through joint training. The method can significantly improve the task completion rate and the emergency response efficiency, optimizes the energy utilization, reduces the system oscillation, and is suitable for disaster monitoring, emergency rescue and other multi-unmanned aerial vehicle cooperative operation scenes.

The application discloses an emergency scene-oriented vertical field large model training and retrieval enhancement method, system, device and storage medium, relates to the technical field of intelligent decision support for emergency scenes, and comprises the following steps: collecting emergency field data and uniformly converting the text, completing segmentation and block division according to semantic breakpoints, and removing duplicates to form a pre-training data set, and constructing a supervised fine-tuning data set on the pre-training data set; after expert sorting and screening according to rules, dividing into reward and punishment models, DPO and PPO training subsets, and selecting a basic model to sequentially perform LoRA continuous pre-training, LoRA supervised fine-tuning and DPO / PPO preference alignment training, and outputting an emergency vertical field large model; constructing a multi-dimensional emergency field benchmark test system through the emergency vertical field large model, and customizing and adapting the emergency vertical field large model according to the functional requirements of different use scenarios. The method disclosed by the application fills the blank of intelligent application in the current emergency field.

This invention relates to the field of industrial equipment condition monitoring technology, specifically to a method for anomaly detection in multi-condition mechanical equipment based on a Gaussian mixture variational autoencoder. The method includes the following steps: synchronously acquiring vibration signals and key phase information; segmenting the vibration signals according to a sliding window and performing fast Fourier transform, logarithmic compression, and sample-by-sample standardization; aggregating the operating condition signals, merging step values, adaptive midpoint boundary discretization, and minimum dwell time jitter reduction; constructing a Gaussian mixture variational autoencoder model based on one-dimensional convolution, and completing health period training through a loss function; establishing a health baseline in the latent space according to operating conditions during the detection phase, and using Mahalanobis distance as the anomaly detection index; and adaptively constructing a threshold using the 3σ principle to achieve anomaly judgment. This method decouples operating condition changes and health degradation without fault labeling, improving the stability and engineering deployability of anomaly detection in multi-condition scenarios.

The invention provides a reservoir water level detection method and system based on periodic segmentation and a dynamic threshold value in the technical field of water conservancy monitoring. The method comprises the steps that S1, water level time sequence data are collected and preprocessed to obtain cleaning data; s2, calculating a dominant period based on the cleaning data; s3, phase segment features are extracted based on the dominant period; s4, mapping the cleaning data to a defined characteristic water level interval based on each phase segment characteristic; s5, respectively calculating a water level change difference sequence for each characteristic water level interval, and calculating an ascending / descending abnormal threshold value based on the water level change difference sequence; s6, inputting the cleaning data into a residual error detection model to obtain a predicted residual error of the water level, and screening an abnormal point from the predicted residual error based on a rising / falling abnormal threshold value; s7, calculating an abnormal confidence score based on each abnormal point; and S8, generating a reservoir water level detection report. The method has the advantages that the accuracy and adaptability of reservoir water level anomaly detection are greatly improved.

This application discloses a method and system for giant panda data detection based on learnable motion saliency modulation, belonging to the fields of endangered animal protection and intelligent infrared image processing technology. It can be applied to the rapid screening of long-term continuous monitoring data from fixed-position infrared cameras. The method constructs a multi-channel temporal difference input for continuous infrared image frames and generates a motion saliency weight map. The weights are embedded as conditional control signals into the convolution calculation process of a lightweight feature backbone network, enabling differentiated feature extraction between motion-saliency regions and static background regions. At the detection output, suppressive modulation is applied to the confidence score to reduce false alarms in static backgrounds. Simultaneously, perturbation frames are identified, and weight updates are maintained or limited to suppress false activations across the entire image and confidence score jitter across consecutive frames. The technical solution of this application solves the problem of balancing efficiency and stability caused by target sparsity, high background ratio, and pseudo-motion interference.

The present application relates to the field of intelligent rail transit communication, and particularly relates to a rail transit communication link anti-interference intelligent switching method, which comprises the following steps: multi-source link data acquisition, link space-time graph modeling, space-time graph neural network interference prediction, multi-constraint switching control modeling, constraint-aware neural network strategy learning and link intelligent switching execution; the present application constructs a space-time graph neural network which fuses time convolution and graph convolution, predicts future evolution of link interference, changes link switching from passive decision based on instantaneous state to active decision based on trend prediction, thereby reducing communication interruption risk and improving foresight of link switching; in the strategy learning stage, time delay constraint, switching overhead constraint and communication reliability constraint are introduced, and a constraint-aware neural network model is formed by embedding a logarithmic barrier function, so that the link switching strategy meets communication quality while taking into account system overhead and stability, thereby improving actual engineering deployability.

The invention discloses a transformer acoustic small sample fault diagnosis method and system based on mahalanobis distance, and the method comprises the steps: simulating a plurality of fault states of a transformer, collecting voiceprint signals in corresponding states as a data set, and dividing the data set into a training set and a verification set; respectively generating corresponding Mel time-frequency diagrams for the voiceprint signals of the training set and the verification set; extracting feature vectors from the training set Mel time-frequency graph and the verification set Mel time-frequency graph by using a convolutional neural network; calculating an average feature vector and a covariance matrix of each fault type based on the feature vectors extracted from the training set; on the basis of each sample in the verification set, the mahalanobis distance between the feature vector of the sample and the distribution of each fault type is obtained, and the fault diagnosis result with the minimum mahalanobis distance is judged to be the fault diagnosis result so as to evaluate the diagnosis accuracy; and for the to-be-diagnosed transformer, acquiring voiceprint signals of the to-be-diagnosed transformer in an operation state, and outputting a fault type according to a Mahalanobis distance minimum principle to complete fault diagnosis. According to the scheme, high-precision fault diagnosis is realized.