21results about How to "Reduce data transfer" patented technology

The invention provides an Internet of Things data acquisition and abnormity early warning method and system, and relates to the technical field of Internet of Things data processing, and the method comprises the steps: constructing a digital twin model for a physical entity at a cloud end, and issuing prediction data to an edge node; acquiring sensor data by an edge computing node at a basic frequency, and computing a residual error between real-time data and predicted data and an information entropy of a residual error sequence; dynamically adjusting the data acquisition frequency of the edge node based on a two-dimensional decision space formed by the residual error and the information entropy; and when the residual error and the information entropy both exceed the threshold values, potential abnormity is judged, an edge side abnormity analysis program is triggered, and abnormal event information is uploaded. According to the invention, through cloud edge cooperation and intelligent decision making, the acquisition frequency is reduced in a stationary period, resources are saved, the frequency is improved when an abnormal symptom appears, monitoring is enhanced, and optimal balance between resource consumption and monitoring precision is realized; double criteria are adopted to effectively filter noise interference, and the false alarm rate and the missing report rate are reduced.

The invention discloses a self-adaptive event triggering cooperative tracking control method for an unmanned cluster system under resource limitation. The method comprises the following steps: establishing a dynamic model of a leader and a follower for a heterogeneous unmanned cluster system; constructing a distributed observer for the follower; determining a state measurement error and an internal dynamic variable of the follower, determining a dynamic event triggering condition of the follower, and constructing a self-adaptive event triggering controller of the follower; when the state measurement error meets or does not meet the dynamic event triggering condition, the adaptive event triggering controller adopts different means to obtain the system observation state of the neighbor node and updates the adaptive event triggering controller; and the follower realizes cooperative tracking control on the leader based on a system observation state. According to the method, the communication burden is effectively reduced, and an efficient and reliable solution is provided for cooperative control of the heterogeneous unmanned cluster system in a complex environment.

This invention relates to the field of flight scheduling technology, specifically providing a method, system, device, and storage medium for collaborative scheduling of low-altitude aircraft. The method includes: edge nodes receiving real-time data from a swarm of drones and generating corresponding scheduling data using a pre-deployed scheduling model; if the edge nodes confirm that the real-time data is environmental mutation data, they upload the environmental mutation data to a regional cloud; the regional cloud matches the environmental mutation data with historical scenario data; if the match is successful, scheduling data is generated; if the match fails, the environmental mutation data is uploaded to a central cloud; the central cloud generates scheduling data based on the received environmental mutation data and returns the scheduling data to the regional cloud, which then distributes the scheduling data to the edge nodes. This invention constructs a low-altitude traffic management system that combines real-time performance, reliability, and self-evolution capabilities.

The invention belongs to the technical field of Internet of Vehicles communication, and discloses a heavy diesel vehicle remote monitoring method and system based on edge computing, and the method comprises the steps: constructing a rule engine according to a predefined transmission rule, and deploying the rule engine to an edge computing node; the data stream of the heavy diesel vehicle is transmitted to an edge computing node, a rule engine is used for evaluating each data field in the data stream, a corresponding transmission level is distributed, and the transmission level comprises a first-level field, a second-level field and a third-level field; and matching a preset transmission strategy according to the data field with the allocated transmission level, executing differentiated data processing, and transmitting a processing result to an enterprise cloud platform through a network transmission layer. According to the method, low-value data is filtered in real time through edge calculation, the invalid data uploading amount is reduced, and data transmission is reduced; in addition, through a high compression ratio and a delay transmission technology, the bandwidth occupation of low-value data is reduced, and the utilization efficiency of network resources is optimized.

The present application relates to a kind of pneumonia detection model training method, specifically to a kind of hierarchical topology pneumonia detection model based on collaborative learning system and method.It overcomes the difficulty of existing pneumonia detection model training system, low efficiency and the problem that local detection node cannot achieve optimal detection performance.First, each node uses local pneumonia data to pre-train respective pneumonia detection model locally, then, center node is based on the feedback data of local model to build hierarchical topology collaborative learning system;After that, each cluster node and child node uses local pneumonia data to train respective pneumonia detection model locally;Finally, collaborative training is carried out, including intra-cluster collaborative training and inter-cluster collaborative training;Intra-cluster collaborative training includes cluster node aggregation training and child node random interactive training.The present application has small data transmission, storage and calculation pressure, and is easy to implement;At the same time, different client is provided with different model parameters, so that each hospital detection node achieves local detection performance maximization.

This invention relates to the field of robot control technology, specifically to an end-to-end closed-loop control method and system for robots based on multimodal perception fusion. It includes: a perception access layer for standardized access to data from multiple sensor types; a fusion processing layer for spatiotemporal alignment and multimodal fusion processing of the standardized access data; a decision control layer for generating execution instructions based on the fusion processing results; and an execution feedback layer for feeding back the execution status to the fusion processing layer to form closed-loop control. This technical solution can solve the problems of standardized access and real-time fusion of multi-source heterogeneous data, and construct a closed-loop control architecture with high fault tolerance and low latency.

This invention discloses a generative image detection method and system for protecting facial privacy. The method first prepares face-text paired data and fine-tunes the diffusion model. Secondly, based on the fine-tuned diffusion model, latent vectors are extracted and clustered, and a dictionary is formed by combining text prompts with the cluster center vectors. Finally, based on the obtained dictionary, pseudo-images and labels are obtained through the fine-tuned diffusion model, and the detection results are output after passing through a classifier. This invention combines latent space clustering with conditional diffusion generation, achieving a balance between privacy protection and data diversity, and significantly improving the security and generalization ability of fake face image detection.

The invention discloses an intelligent report generation system and method based on knowledge graph enhancement, and belongs to the technical field of natural language processing and knowledge graph fusion. In order to solve the problems of insufficient accuracy, low efficiency and opaque reasoning process in the existing structured report generation only depending on large model intrinsic knowledge, the method comprises the following steps of: constructing a knowledge graph containing entities, attributes and relationships, and converting a user natural language problem into a structured query statement by utilizing a first large model; after an entity relationship result is obtained in the knowledge graph, a report with a clear basis is generated by the second large model; and adaptation can also be realized through query languages. By means of structured knowledge and efficient query of the knowledge graph, the accuracy and generation efficiency of the report conclusion are improved, the reasoning process is traceable, meanwhile, the flexibility and universality of the scheme are enhanced, and the method is suitable for structured report generation scenes such as enterprise data analysis and industry research.

The application discloses a task scheduling method for general applications of a heterogeneous integrated chip system. The method aims at the problem that the computing capacity, on-chip storage capacity and cross-grain communication overhead of different grains in the heterogeneous integrated chip are quite different, constructs a system architecture graph and an operator dependency graph, and establishes unified computing delay, communication delay, storage constraint, cost and power consumption models. On this basis, the task mapping process is modeled as a Markov decision process, a local subgraph sampling, a communication-aware graph attention network and a cross-attention mechanism are used to encode the scheduling state, and a proximal policy optimization algorithm is used to search the mapping scheme of the operator to the processor core. The method can consider the computing load balancing, on-chip storage feasibility and cross-grain communication cost under the premise of meeting the resource and constraint conditions, and is suitable for application scheduling of large model inference, graph computing, deep learning training and inference and other applications with complex dependency relationships.

The invention belongs to the technical field of enterprise data management, and provides an artificial intelligence-based enterprise data management method and system, and the method comprises the steps: carrying out the comparative analysis of incremental data and total data in a data fusion processing process of the incremental data and the total data after the collection of enterprise big data, and obtaining a data fusion result; whether a Cartesian product appears or not is recognized, when the Cartesian product appears in the data fusion processing process, the impact amplitude of the Cartesian product on data fusion processing is evaluated according to the fusion data expansion degree and the fusion data coincidence degree, and it is guaranteed that key information is not missed in the data fusion process by evaluating the impact amplitude; when the impact amplitude is within a controllable range, effective integration of total data and incremental data can be ensured, an enterprise can have a complete data view, the pressure of data fusion on system resources is relieved, system crash or performance reduction caused by too large data size is avoided, and rapid data query and processing capacity is provided for the enterprise.

The invention provides a data reconstruction method, a distributed storage system and a computer readable storage medium, and is applied to the technical field of distributed storage. The data reconstruction method is applied to a proxy repair service, the proxy repair service is deployed at a persistence layer client of a distributed storage cluster, the distributed storage cluster comprises a plurality of nodes, and each node is configured with the persistence layer client and an object storage server. According to the method, the proxy repair service is applied in the distributed storage cluster, local loopback communication between the object storage server and the persistence layer client is realized, and cross-node network data transmission is remarkably reduced by reading local source fragmented data and executing erasure code reconstruction, so that the data reconstruction network bottleneck in a large-scale EC scene is effectively relieved, and the data reconstruction efficiency is improved. And the data reconstruction efficiency and the overall availability of the cluster are improved.

The embodiment of the invention relates to the technical field of computers, and provides a data acquisition method and electronic equipment, a first application and a first test tool are installed on the electronic equipment, and the first application comprises a reflection proxy service and a reflection interface. The electronic device determines a first reflection parameter through a reflection proxy service in response to a first data request of a first test tool. Wherein the first data request is used for requesting to obtain feature data of the first object, and the first reflection parameter comprises an identifier of the first object. Then, the electronic equipment transmits the first reflection parameter to the reflection interface through the reflection proxy service, the reflection interface is called to obtain first data generated by operation of the first application, and the first data comprises feature data of the first object; finally, the electronic device returns the first data to the first test tool through the reflective proxy service. Through the method, the complexity of cross-process communication interaction can be effectively reduced, and the data acquisition efficiency is improved.

The application discloses a proportional-integral synchronization controller construction method and device for a complex dynamic network, and the method is as follows: a mathematical model of the complex dynamic network with a Markov switching topology, a dynamic event triggering mechanism and a proportional-integral control strategy is established; an exponential synchronization target in the complex dynamic network in the mean square sense is designed; a Lyapunov function is constructed, and a sufficient condition for the exponential synchronization of the system in the mean square sense is deduced; by solving a linear matrix inequality, a gain matrix of the proportional-integral controller required to ensure the synchronization performance of the system and parameters of an event triggering condition are designed; Zeno behavior is excluded, and the effectiveness of the dynamic event triggering mechanism is verified. The device comprises the following modules: a mathematical model establishment module, an exponential synchronization target design module, a condition deduction module, a proportional-integral controller design module and an effectiveness verification module. The application improves the stability performance of the complex dynamic network system, reduces the steady-state error of the system, and has the advantages of simplicity, easy realization and strong applicability.

This invention discloses a method, apparatus, device, and medium for updating a fundus image recognition model, relating to the field of ophthalmology. The method includes: determining a set of target features for the fundus image based on the acquired knowledge of fundus image recognition model samples corresponding to the user terminal to be updated; acquiring the activation values ​​of neurons in the recording network corresponding to each user terminal to be updated; constructing a local fundus image recognition model subgraph based on the fundus image recognition model sample knowledge, activation values, and a set of common features of the fundus image; transmitting all local fundus image recognition model subgraphs to a server; performing subgraph fusion on all the local fundus image recognition model subgraphs to obtain a fused subgraph; and updating the local fundus image recognition model using the fused subgraph. This invention transforms the fusion and updating of the user terminal model into an operation of subgraph fusion to generate a full image, ensuring privacy and security throughout the model fusion cycle based on multiple privacy computing technologies, and achieving secure model transmission and incremental fusion.

The present application relates to the technical field of equipment state monitoring, and more particularly to a method and system for detecting bolt loosening of an emergency shut-off valve based on multi-modal fusion, specifically comprising the following steps: multi-source data synchronous acquisition, accelerometer installation and data processing, environmental adaptability preprocessing, bolt region optical flow feature extraction, Kalman filter multi-source fusion, bolt loosening detection and loosened bolt positioning. The method for detecting bolt loosening of an emergency shut-off valve based on multi-modal fusion realizes online, automatic and real-time accurate monitoring and early warning, significantly improves the reliability and anti-interference ability of the system, provides high-precision and quantifiable objective diagnostic results, has good economy and large-scale deployment feasibility, and effectively solves the industry pain point that traditional contact-type sensor solutions are difficult to be applied in large-scale in industries with a large number of bolts due to high cost and technical complexity.

This invention relates to the field of network technology and discloses a method, computer device, storage medium, program product, and network system for determining the location information of routing nodes. The method includes: acquiring a first code and a second code of a first routing node, and acquiring a third code and a fourth code of a target routing node; determining whether the first routing node and the target routing node belong to the same network subsystem based on the second code and the fourth code; when it is determined that the first routing node and the target routing node belong to the same network subsystem, performing an XOR operation on the first code and the third code to obtain a target code; determining the number of target values ​​included in the target code; and determining the location information of the next routing node corresponding to the first routing node based on the number of target values, the first code, the second code, the third code, and one or more elements in the target code. This invention can improve network transmission efficiency.

The invention discloses a connecting device for an escalator and a control method, the connecting device comprises double physical safety links, and two completely independent and electrically isolated physical safety links, namely a link A and a link B, are laid in parallel on the whole line of the escalator; the distributed safety node comprises a safety relay unit which comprises two pairs of normally open contacts KA and KB which are forced to guide and are completely synchronous in state, and the two pairs of normally open contacts KA and KB are respectively connected in a physical safety link A and a physical safety link B in series; the physical link on-off detection circuit is used for directly detecting voltage / current of a link A and a link B and judging the on-state of the link A and the link B, and the diagnosis excitation and measurement circuit comprises a precise constant current source and a high-precision voltage sampling and demodulation circuit and can be switched to a diagnosis bus. A test signal is injected into a diagnosis loop formed by a plurality of node coils, and a response is measured, so that a tough and adaptive safety protection system is constructed, and potential network attacks and internal faults can be resisted.

This invention relates to the field of rehabilitation data communication technology, specifically a method for integrated medical and educational multimodal data acquisition and analysis for autism rehabilitation. The method includes: constructing a set of multimodal data acquisition units; performing integrated autism rehabilitation data detection based on these units to obtain multimodal rehabilitation data; if the multimodal rehabilitation data contains anomaly tags, transmitting the data to the original receiving terminal via an emergency transmission channel to obtain the target receiving terminal; if the multimodal rehabilitation data does not contain anomaly tags, recording it as normal polling data; summarizing the normal polling data to obtain a normal polling dataset; and transmitting the normal polling dataset to the original receiving terminal to obtain the target receiving terminal. This invention can improve the accuracy of rehabilitation status monitoring and reduce the risk of missed reports of abnormal conditions.

The embodiment of the invention provides a coding method, a hardware accelerator and a system on chip, and the method comprises the steps: determining a cardinal number frequency corresponding to each to-be-coded cardinal number obtained by mapping each input symbol through a frequency extraction module in the hardware accelerator; generating a target Huffman tree through a code generation module in the hardware accelerator based on the target sorting result of the cardinal number frequency corresponding to each cardinal number to be coded; and generating a compressed data stream through an encoding module in the hardware accelerator based on a target encoding result obtained by dynamically encoding each symbol to be encoded in the input data stream. And constructing a Huffman tree structure according to the target sorting result, so that the high-frequency cardinal number corresponds to the short code and the low-frequency cardinal number corresponds to the long code. And dynamically coding the input data stream based on the target Huffman tree to generate a compressed data stream, thereby realizing rapid coding conversion. The whole coding process is efficiently and cooperatively completed in a hardware accelerator, so that the data transmission and software scheduling overhead is reduced, and the data compression speed and performance are improved.

The application discloses an indoor space lighting fast response system based on edge computing, which comprises a space state acquisition module, an edge computing processing module, a mode determination module, a mode consistency constraint module, a parameter updating module and a lighting control output module.