35results about How to "Guaranteed data privacy" patented technology

The invention discloses a network security threat monitoring method and system based on artificial intelligence, and relates to the technical field of digital information transmission, and the method comprises the steps: obtaining data of a plurality of data sources in a network environment, carrying out the multi-level data fusion of the data through an artificial intelligence model, and forming unified data representation; the artificial intelligence model comprises at least one selective fusion layer, and features are extracted from different data sources and are jointly expressed; based on a data fusion result, performing anomaly detection by adopting an unsupervised learning method; and modeling the network flow data into a graph structure according to the detected abnormal behavior, and analyzing the graph structure by adopting a graph neural network GNN. According to the method, the technical problems of difficulty in multi-dimensional data fusion, delay in threat detection, insufficient response capability and the like in the existing network security protection technology are solved.

The invention discloses a shield tunnel risk early warning method and system based on digital twinning and federated learning, and the method comprises the steps: a shield tunnel construction side client serves as an edge calculation terminal, and obtains basic parameters; a first shield digital twinning system, a second shield digital twinning system and a third shield digital twinning system are correspondingly constructed by the shield machine producer client side, the shield tunnel construction side client side and the third-party client side according to the basic parameters; a shield construction risk early warning model is arranged in each shield digital twin system; the shield machine production side client, the shield tunnel construction side client and the third side client are all participants, and each participant sends model parameters of the respective shield construction risk early warning model to the central server; the central server is used for averaging the parameters of the three models and issuing the average value of the parameters to each participant; and each participant updates the model of the participant according to the average value of the parameters, and realizes shield construction risk early warning based on the updated model.

The invention provides an abnormal network traffic identification method based on small sample data enhancement, and the method comprises the steps: S1, converting one-dimensional traffic data into a two-dimensional gray-scale map, and employing a twin network combined with a dynamic attention mechanism to analyze the differences between various types of traffic data pair by pair; parameters of an attention mechanism in the twinborn attention network are analyzed, and inter-class difference features are identified; s2, selecting a training domain by using the inter-class feature difference matrix, introducing a comparative learning method to judge the difference between real data and generated data, and generating an enhanced data set for each class of traffic; s3, after the enhanced data set issued by the central server is received, each edge server uses the enhanced data set and local sample data to train the twinborn attention network in the step S1; due to different local sample data sets, each edge server is trained to obtain a personalized abnormal traffic identification and classification model suitable for local data distribution, and abnormal traffic is effectively classified.

The invention discloses a medical auxiliary diagnosis method and electronic equipment based on quantum federated learning in the technical field of smart medical, and the method comprises the steps: carrying out the medical auxiliary diagnosis based on a pre-trained quantum federated learning model, and the obtaining process of the quantum federated learning model comprises the steps: constructing the quantum federated learning model, comprising a local client and a central server. Repeatedly executing the following steps until a preset training period is reached: in the current training period, training the local client according to the training sample set, and calculating the accuracy of the local client according to the test sample set; the local client sends the accuracy and the final model gradient to the central server after the current training period; the central server updates global model parameters according to the accuracy and the model gradient, and returns the global model parameters to the local client; and the local client updates the local model parameters according to the global model parameters. According to the invention, the accuracy and efficiency of medical auxiliary diagnosis can be improved.

The invention belongs to the technical field of data processing. The invention provides a spatio-temporal data prediction method and system for heterogeneous photovoltaic diagram structure data fusion, and the method comprises the steps: a current client obtains a photovoltaic power generation prediction result of the current client at each moment in the future after a current moment based on the historical photovoltaic power generation data of the current client and a trained global model; the global model training process comprises the steps that the central server distributes a global model pre-trained according to a photovoltaic generating capacity data set to all clients, and each client trains the global model through independent local graph data to perform local updating; each client uploads a local update result to a central server for aggregation, and the central server trains a GCN part of a global model according to an aggregation result and distributes the trained global model to each client for use; on the premise of ensuring data privacy of each client, updating and training of a global model are realized, and prediction comprehensiveness is greatly improved.

The invention belongs to the related technical field of image classification, and discloses a continuous image classification model training method and system for realizing zero sample playback and simple extension, and the training method comprises the steps: solving an optimization objective function facing a current classification task unit base vector; distributing an optimal equal-included-angle base vector matrix of the current classification task from a unit hypersphere, and adding the optimal equal-included-angle base vector matrix into an equal-included-angle base vector matrix set which stores equal-included-angle base vector matrixes of historical classification tasks to update the set; in the forward propagation of the training, the optimal global output representation of each hidden layer is determined by solving a minimum statistical dependence problem; and in back propagation of training, network parameters of each hidden layer are fed back and adjusted by using a gradient descent method. According to the method, the balance between the stability and the plasticity can be realized on the continuous image classification task, the method has the characteristics of zero sample playback and simple expansion, and the training efficiency is greatly improved by improving the optimization problem of each step.

The invention relates to the field of electric power, in particular to a source network load storage multi-source heterogeneous data fusion method based on cross-feature federated learning, comprising the following steps: S1, collecting participant data, and preprocessing the participant data, the preprocessing comprising heterogeneous data feature extraction, secret generation, shared encryption, and model training and evaluation; performing data fusion processing on the preprocessed participant data to obtain fused participant data; and S2, carrying out decryption recovery based on the processed participant data to obtain updated global model parameters, and sending the updated global model parameters to the participant to generate new participant data for next iteration. The multi-source heterogeneous data fusion method has the advantages that the multi-source heterogeneous data can be safely fused on the premise of not leaking the privacy data of each participant by performing feature extraction on the heterogeneous data and based on the privacy computing technologies such as homomorphic encryption and secret sharing, and the data security is effectively improved.

The invention discloses a cloud desktop throughput prediction method for cold start and mode forgetting. The method comprises the steps of 1, learning aided training based on similarity attention federal; 2, throughput prediction for safe cold start in the initial stage of a new user is carried out; step 3, fluctuation-residual feature extraction based on a TDLinear network model is carried out; 4, updating the model based on the adaptive memory module and the EWC; and step 5, training forgotten throughput prediction in the long-term use stage of the user. Through a federated learning algorithm, on the premise of ensuring user data privacy, existing throughput historical use knowledge of other users is utilized to assist training. An old mode is continuously reviewed based on continuous incremental prediction, a memory module is designed to incrementally update a new mode, and the method has the advantages of avoiding forgetting in training and integrating two stages to realize full-life-cycle throughput prediction of the cloud desktop user.

This invention discloses an intelligent empowerment system for automotive after-sales service and its implementation method. The system includes a case collection module, a case review module, a knowledge sharing module, an intelligent recommendation module, a mutual assistance and communication module, a backend data analysis module, and a business empowerment module. Dealer repair personnel can submit standardized fault cases to the platform. After review, a case database is formed. The system calculates recommendation weights based on user profiles, case quality, and interaction data, achieving personalized distribution and precise matching of cases. Users can browse cases, seek help, and engage in interactive learning within the platform, forming a technical mutual assistance community. The backend data analysis module structurally extracts the collected cases and interaction information, generating standardized technical intelligence, and pushes it to business systems such as technical support, warranty management, spare parts planning, and training management, building a two-way empowerment mechanism between OEMs and dealers, and achieving closed-loop management from fault cases to business optimization.

The invention provides a multi-party security intersection solving method, device and equipment and a storage medium, and belongs to the field of information security, and the method comprises the steps: obtaining a first data set and a first secret key of an intersection solving party, and encrypting a data ID in the first data set according to the first secret key to obtain a first ciphertext; sending the first ciphertext to a first data provider, so that the first data provider encrypts the second ciphertext based on the first ciphertext to obtain a third ciphertext; obtaining a third ciphertext, and performing data intersection pre-screening on the first data set according to the third ciphertext and the first key to obtain a first candidate data set; acquiring a second candidate data set sent by the first data provider; and performing data intersection processing on the first candidate data set and the second candidate data set to obtain a target data set existing in all data parties. According to the method, the data privacy of each data party is ensured, the data intersection of multiple parties can be obtained, and the accuracy and the security of the data intersection are greatly improved.

This invention provides a network attack detection method and system for vehicle-to-grid (V2G) interaction. It includes: a local client capturing raw traffic data and extracting key features; then receiving the latest global model parameters, detecting the raw traffic data and encoding the attack detection results into metadata tags, while simultaneously using the key features to train the global model parameters to calculate the quantization gradient; extracting feature vectors from the raw traffic data and combining them with the metadata tags to perform attack prediction and obtain the attack prediction probability; the local client executing defensive actions based on reinforcement proxy learning on the attack prediction probability; the local client uploading the quantization gradient and metadata tags to the server via a secure sparse aggregation protocol; the server reconstructing the global gradient to update the global model parameters via a key exchange protocol and updating the attention weights of the global model using the metadata tags; the latest global model outputting the attack probability of V2G interaction, determining whether to issue an alarm and execute defensive actions.

The invention provides an intelligent breeding behavior analysis system and method based on edge calculation and AI vision, relates to the technical field of intelligent breeding and computer vision, and solves the technical problems of low efficiency, poor behavior recognition accuracy and abnormal early warning lag caused by dependence on manual and cloud processing in existing breeding management. The system comprises an edge end device which is used for collecting a video stream in real time and carrying out target detection and key point identification on the video stream through a lightweight deep learning model to obtain a preliminary behavior classification result; the fog end server is used for gathering a plurality of preliminary behavior classification results, performing complex behavior recognition, behavior time sequence analysis and abnormal behavior detection, obtaining an abnormal behavior result and triggering real-time early warning; and the cloud platform is used for receiving and storing all data of the farm, performing long-term data mining, multi-source data association analysis and model training and updating, and generating decision support information. The method is used in the intelligent breeding process.

The invention discloses an inter-cloud cross-domain cooperation method and system for controlled access of global medical data. The method comprises the following steps: a node A creates a transaction request according to an access attribute, carries out uplink and appoints a node B as a receiver; if yes, the node B generates an access token containing restriction information, binds the access token to the node A to generate transaction confirmation, and carries out uplink and appoints the node A as a receiver; the node A takes the access token as a credential for accessing data in the node B, and sends a data access request to the node B through a specified interface; and the node B verifies the access token carried in the data access request and controls whether the data access request is allowed or not. The invention aims to realize sharing and controlled access of global medical data in a cloud environment, eliminate data barriers among private clouds of different hospitals and promote data intercommunication and cooperation among different medical institutions, so that the need of repeated medical examination is reduced, the total data storage space is reduced, and the overall data utilization rate is improved.

The invention discloses a virtual power plant aggregation operation cost scheduling method based on an input convex neural network, and the method comprises the steps: constructing a feature coding model based on a deep neural network, carrying out the feature extraction of the output data of a distributed power supply in a virtual power plant, and obtaining the coding feature data; constructing an equivalent transformation model of the input convex neural network, and solving to obtain a multi-period aggregation operation cost agent model; and taking the obtained multi-period aggregation operation cost agent model as an operation cost function of the virtual power plant, constructing a multi-period random economic optimization scheduling model of the power distribution network, solving the model, obtaining an optimal aggregation power scheduling instruction of the virtual power plant, and issuing the optimal aggregation power scheduling instruction to execute scheduling. According to the method, implicit feature coding is utilized to guarantee data privacy, a nonlinear relation is accurately fitted by inputting the convex neural network, mathematical convexity is strictly guaranteed, the difficulty of non-convex solution is overcome, and it is guaranteed that a multi-period scheduling scheme has global economic optimality.

The present invention relates to a method, device, electronic device and storage medium for cross-domain tracing of an off-chain oracle. The cross-domain tracing method of an off-chain oracle includes the following steps: data registration, in which an external data provider registers the data source it provides to an off-chain oracle, and the external data provider is a third-party service and system; data request, in which if the business system uses a different chain, the smart contract sends a data request to the off-chain oracle, and synchronizes the contract data through cross-chain technology; data acquisition, in which after the off-chain oracle receives the data request, it obtains the corresponding data from the registered data source; data transmission, in which the off-chain oracle transmits the acquired external data back to the smart contract; data processing, in which after the smart contract receives the external data, it processes and verifies the data in the contract. According to the cross-domain tracing method of an off-chain oracle of the present invention, it is possible to realize off-chain traceability audit applications, use the oracle contract as a data source, conduct traceability queries, and adapt to the business needs of multiple parties.

The invention relates to the technical field of material management, and discloses a material management system based on a data base, comprising a business platform, a data base and a basic module which are in communication connection in sequence. According to the invention, the problems of errors of statistical data of each application module and the like in the prior art are solved.

The invention provides a data sharing method of electronic equipment and a related device. The method comprises the following steps: receiving a click operation on a virtual icon corresponding to a memory data viewing function in a main interface of first electronic equipment; the first electronic device is controlled to start a camera and scan a target entity photo to obtain target authority verification data, and the target entity photo is associated with target data content uploaded to the cloud server by the second electronic device; sending a permission verification request to a cloud server, wherein the permission verification request comprises target permission verification data and equipment information of the first electronic equipment; receiving target data content associated with the target entity photo sent by the cloud server, wherein the target data content is sent after the cloud server confirms that the first electronic equipment has the permission of looking up the target data content based on the permission verification request; and controlling the first electronic equipment to output the target data content. According to the invention, data sharing of the electronic equipment is realized on the premise of ensuring data privacy.

The invention belongs to the field of WiFi perception and emotion recognition, provides an emotion perception system and method based on WiFi channel state analysis, solves the problem that an existing system is high in cost, and facilitates laying of the system. The system adopts mobile equipment as a signal processing module, so that the system is more convenient to use; the judgment model uses the relationship between the human body breathing mode and the human emotion, and the limitation of the existing emotion perception system based on the human body action posture is improved; in addition, the system also adopts a multi-channel contrastive analysis method to eliminate sudden factors during signal acquisition as much as possible, so that the emotion judgment result of the system is more pleasant, and the anti-interference performance of the system is improved; and a database updating mechanism and a privacy security mechanism are also added, so that the data privacy of the user is ensured, the database is continuously expanded along with the increase of the user and the feedback of the user on the judgment result, and the data collection scene is continuously enriched, thereby enabling the judgment accuracy of the system to be higher and higher.

The application discloses a large language model driven intelligent financial field database interaction method, and relates to the field of databases, and the method comprises the following steps: S1, constructing a language model and training; S2, recognizing and analyzing user input information by using the language model; S3, the language model selects a database interaction tool according to the recognition result, and queries the database; S4, the database interaction tool returns the query result, and saves the query data to an intermediate file; S5, steps S3-S4 are repeatedly executed based on an action chain until all database interaction tools finish querying, and the query result is output through the language model; S6, the output variable is replaced by using the query data in the intermediate file, and the result is returned to the user. The application reduces the threshold of user and database interaction, and realizes the innovation of user interaction.

The present invention discloses a trusted federated learning system and method based on a trusted execution environment, including a server and several clients; the server and the several clients are all configured with a domain virtual machine, an encryption module and a decryption module, and the domain virtual machine is used as a trusted execution environment, in which a remote secure connection module is configured. The present invention protects the local training and central aggregation process of the entire federated learning based on the trusted execution environment, and the global model and the local model and their updated data only exist in plain text in the trusted execution environment. The data is encrypted when leaving the trusted execution environment, protecting the integrity of the federated learning protocol and the confidentiality of the parameters, effectively resisting poisoning attacks and privacy attacks while only introducing very small computing overhead, effectively protecting user data privacy while improving the accuracy and security of the trained model, and comprehensively solving the privacy and security attacks faced by the current federated learning system.

The present application relates to a kind of ocean ranching intelligent decision system based on self game, comprising: multi-modal data perception module, for collecting multi-modal data and constructing space-time correlation feature atlas by graph neural network;Large model decision hub module, for mapping multi-modal data to unified semantic space by cross-modal alignment technology, and variable correlation reasoning based on injected marine ecological causal diagram;Dynamic game optimization module, for taking ecological protection target and economic benefit target as game parties, and searching Pareto optimal solution;Virtual-real linkage execution module, for driving ocean ranching digital twin and physical equipment to perform corresponding operation according to decision instruction, and realizing dynamic updating of model parameters by federated learning.The present application can improve the yield of cultivation, reduce ecological disaster loss, while supporting carbon sink asset digitalization certification, providing ocean ranching with intelligent decision support with ecological adaptability and economic robustness.

The application discloses a kind of group learning incentive methods based on blockchain, first the node in cluster is divided into network node, data node, computing node with incentive mechanism and other nodes, the incentive method based on blockchain and reputation mechanism is used to main node, it is supplemented with the reward and punishment mechanism based on game theory to calculate the possible reward and fine of each node, make the benefit of each node honest work maximum.The application can realize the virtuous cycle of group learning, rely on the reputation recorded by blockchain and game theory method to promote data node to provide real information;In the era of deep learning and machine learning, it can play an important role, such as between various financial institutions, this cannot make them trustingly carry out privacy protection data fusion, but the application can evaluate the credit between each institution, maintain the benefit balance of data demander and data provider, so that each party voluntarily joins the group learning process, and fully develops potential data value.

The invention relates to a multi-station wind power prediction method based on feature embedding-transform federated learning. The method comprises the following steps of 1, data preprocessing, 2, feature embedding, 3, transform prediction model construction and 4, federated learning joint training. According to the scheme, a dynamic feature embedding algorithm is introduced, feature embedding calculation is carried out on current features before the features are input into a model, on one hand, the problem that feature data of different stations are inconsistent is solved, and on the other hand, the correlation between different features can be fully utilized through introduction of a dynamic word embedding algorithm. The transformer can directly capture the dependency relationship between any two positions, is not influenced by the sequence length, and can better process the prediction problem of the time sequence data, and in addition, the transformer can completely realize parallel calculation relative to the lstm, so that the calculation efficiency of the model is greatly improved.

The invention belongs to the technical field of intelligent charging pile cooperative scheduling, and discloses a new energy automobile intelligent charging pile cooperative scheduling system based on the Internet. The system comprises a user charging demand prediction module, a virtual power grid demand aggregation module, a predicted charging demand charging scheduling module and an automatic cost settlement transfer module, user charging demands are predicted based on an overall prediction model, predicted charging demand reports are obtained, all the predicted charging demand reports are integrated, and the integrated charging demand reports are obtained. Based on the power grid state data, an optimization target report is generated, a reinforcement learning algorithm is created for each predicted charging demand, an optimal charging scheduling report is generated for each predicted charging demand, and automatic cost settlement and transfer are performed based on event data in the block chain. The method has the remarkable advantages that the user charging demand prediction capability is high, the global scheduling optimization effect in the region is good, and the vehicle-power bidirectional mode application effect is great.

The invention discloses a privacy auditing single-round federal learning method and device for biomedical data, and relates to the technical field of artificial intelligence and privacy protection, and the method comprises the steps: carrying out initialization, generating or receiving a CKKS public and private key pair and an ECDSA signature key pair by a medical institution, and initializing an auditing threshold by a central research institution; each medical institution trains a local model based on a local training set to output and obtain features, encrypts and signs the output features, and uploads the features to the central research institution; and the central research institution performs signature verification and ciphertext analysis so as to perform encrypted domain data auditing and verification, and realizes encrypted domain security aggregation and knowledge distillation in combination with threshold setting and self-adaptive updating. According to the application, end-to-end encryption transmission, uploaded data identity verification and model quality auditing in an encryption domain under biomedicine multi-center cooperation can be realized, and data privacy, model authenticity and result traceability are ensured.

The invention belongs to the technical field of network security, and discloses an intelligent label construction and rapid search method for distributed data privacy protection. According to the method, the correlation interpolation function is constructed, the randomization of the main-sub label mapping is ensured, the scale of the number of function items is controlled, and the main-sub label mapping is segmented and stored in the two servers in an addition secret sharing form, so that the safety of label mapping storage is improved, and the balance between the safety and efficient storage is achieved. Besides, the search algorithm converts the query problem of the sub-tags into a function value solving problem, the tedious process of data comparison in the query process is avoided, the higher query speed can be obtained, and the requirement for large-scale data split storage can be met.

The invention discloses a client and sample joint optimization method for federated learning of the industrial Internet of Things, which comprises the following steps: S1, constructing a cloud-side-end architecture, S2, initializing a cloud server, S3, executing importance perception data shuffling by terminal equipment, S4, screening clients participating in federated learning training by an edge server, and S4, carrying out data shuffling by the edge server. S7, the edge server executes synchronous aggregation and uploads the updated data to the cloud server; S8, the cloud server executes global aggregation and issues the updated data to the edge server for synchronization; S9, the steps S3-S8 are repeated, and S7, the selected client device executes grouping importance sampling; S7, local model training is executed and the updated data is uploaded to the edge server; and the model converges or reaches a preset training round. According to the method, the training efficiency, the convergence speed and the prediction precision can be remarkably improved in a resource constraint scene.

The invention discloses a tonsil postoperative diet management system and method, and relates to the technical field of medical care, and the system comprises a front end, a rear end, a database, an image recognition module and an intelligent interaction module. All the parts collaboratively realize the following functions: a data management function, a system management function, an authority management function, an automatic recipe recommendation function, an image recognition function and an interactive forum function. The personalized diet recipe can be automatically recommended according to the basic information, the medical history, the number of postoperative days and the tonsil mucosa recovery condition of the patient, meanwhile, an interaction platform between a doctor and the patient is built, and the problems that existing clinical diet guidance is general, individuation is lacked and doctor-patient communication is inconvenient are solved.

The present application relates to the technical field of transformer cooling, and discloses a transformer adaptive spraying cooling method, system, device and medium. The method comprises the following steps: based on the multi-dimensional working condition data of each main transformer, an oil temperature prediction value of the next moment of the corresponding main transformer is determined by using a machine learning model combining federated learning and gradient boosting machine; a first oil temperature deviation between the oil temperature prediction value and a preset target oil temperature and spraying energy consumption are used to construct a flow control target function corresponding to each main transformer, each flow control target function is solved according to the corresponding flow control constraint condition, and the optimal nozzle flow of the corresponding main transformer is obtained; and based on the optimal nozzle flow of each main transformer, the output power of the water pump of the corresponding main transformer is updated in real time, so that adaptive spraying cooling is performed on each main transformer. The present application realizes the precision, energy saving and safety of transformer spraying cooling.

The invention relates to a multi-client collaborative SAR image target recognition method based on a semi-decentralized architecture, and the method comprises the steps: collecting SAR image data, carrying out the preprocessing, generating a corresponding target recognition tag, and obtaining a data set with a tag; building a semi-decentralized federated learning framework, and initializing global model parameters; each client performs model training by using local private data, and obtains a local parameter consensus through communication between devices; the server randomly selects the clients for parameter transmission, aggregates model parameters, updates global model parameters, and broadcasts the updated global model parameters to all the clients; iterative training is carried out until global model parameters reach a convergence state, and a trained global model is obtained; and testing the trained global model, and selecting the tested global model to carry out SAR image target identification. Compared with the prior art, on the premise of ensuring model performance and data privacy protection, the communication overhead between the server and the client can be effectively reduced.