1488 results about "Multiple Models" patented technology

A system and method for flexibly incorporating machine learning models into applications using a marketplace platform and distributed computational graph (DCG) architecture. The DCG enables dynamic selection, creation and incorporation of trained models with data sources and marketplaces for data, algorithms, simulation models, ontologies, knowledge corpora, and crowd or expert judgment. Multiple models can be used in series or parallel. An expert judgment marketplace allows human and artificial intelligence (AI) experts to score the accuracy of training data and model outputs. Consumers can select and rank AI agents or experts based on the helpfulness of their judgments. A symbolic knowledge corpora and retrieval augmented generation (RAG) marketplace enables selling access to proprietary datasets as RAGs and knowledge bases. The system includes knowledge corpora and RAG marketplaces with domain-specific components and user experience customization.

The invention discloses a financial data risk control system and method based on big data, and relates to the technical field of financial science and technology, and the system comprises a multi-source data collection module which achieves cross-mechanism safety collection through federal learning; the data cleaning and preprocessing module is used for processing abnormal values and missing values by using an improved algorithm; the knowledge graph construction module is used for constructing a dynamic knowledge network based on an innovative algorithm; the risk assessment engine fuses various models to assess risks; the real-time monitoring and early warning module is used for realizing second-level response by utilizing multi-scale analysis; the decision support module is used for optimizing a strategy based on reinforcement learning; and the audit tracking module guarantees evidence storage and privacy through zero-knowledge proof, and all the modules cooperate to improve the risk control capability. According to the financial data risk control system and method, risks are accurately recognized, real-time monitoring and early warning are achieved, data security sharing is achieved, risk control strategies are dynamically optimized, risks and business development are balanced, the risk prevention and control capacity and economic benefits of financial institutions are improved, and data privacy and risk control transparency are guaranteed.

The invention discloses a collaborative scheduling method based on a security agent, and relates to the technical field of network security design. The specific operation of the security agent collaborative scheduling method comprises the steps of system deployment and initialization, security operation task execution process, agent self-learning and capability evolution implementation, cross-domain security agent collaborative adaptation implementation and security operation visualization and traceability implementation. Connection paths between the security agent and business data, a private knowledge base, a security tool and multiple models are broken through, the problem that all elements in a traditional mode lack efficient communication and collaboration is solved, and by means of agent routing dynamic scheduling, A2A protocol interaction, MCP protocol tool calling, RAG business interface calling and multi-model combination, the security of the security agent is improved. Integration and intellectualization of the safety operation process are achieved, the continuity and the response speed of the operation process are improved, safety operation is more efficient and collaborative, and complex and variable safety requirements are met.

The invention relates to the technical field of Internet of Things, and provides a wire and cable online quality detection method and device, electronic equipment and a storage medium. The method comprises the following steps: performing multi-scale wavelet packet decomposition and signal entropy fusion on an original wire and cable signal matrix to obtain a feature tensor, and performing anomaly recognition extraction on the feature tensor through a graph attention network model to obtain a potential defect feature vector; and performing multi-physics coupling simulation inversion on the potential defect feature vector to obtain a defect quantization parameter set, performing dynamic reasoning according to the defect quantization parameter set through a Bayesian network model to obtain an online quality grade decision, and packaging the online quality grade decision according to a block chain intelligent protocol to obtain a quality traceability record. According to the invention, through organic coupling of multiple levels, multiple models and multiple technical means, real-time performance, accuracy, traceability and safety of online quality detection of wires and cables are realized.

The invention discloses a cloud data anomaly detection and safety response system based on artificial intelligence, relates to the technical field of data processing, and solves the problems that firstly, an incremental compression algorithm is difficult to store and preprocess multi-source heterogeneous data; secondly, it is difficult to fuse statistical analysis and a deep learning model, locate outliers and analyze abnormal semantics in unstructured data, and then it is difficult to effectively predict a potential attack path; then, on the premise that data security and traceability are guaranteed, correlation analysis of cross-node anomalies is difficult to achieve so as to identify distributed attacks; and finally, an attack and defense confrontation model is difficult to construct for safety response, and the safety response effect is difficult to evaluate. According to the method, cloud environment data are processed through an adaptive probe cluster and the like, multiple models are fused to generate anomaly detection features and predict attack paths, and response and evaluation are carried out through reinforcement learning and digital twinning by means of cooperative detection such as federated learning and the like.

The invention provides a weather forecast learning system based on an artificial intelligence algorithm, and the system comprises a data access collection module which collects a multi-source data set; the data fusion and processing module is used for carrying out data fusion and processing to obtain a multi-source fusion data set; the AI model design module is used for constructing a multi-model collaborative architecture and carrying out multi-model parallel training and optimization; the model fusion module is used for carrying out multi-model weighted fusion to obtain an AI model; the real-time prediction module is used for updating a prediction result according to the real-time data; the visualization and interpretability module is used for designing a visualization interface and carrying out interpretability verification; and the evaluation and iteration module is used for carrying out comprehensive evaluation and continuous improvement on a prediction result in combination with evaluation indexes. According to the method, accurate and reliable observation data can be obtained, massive meteorological data are efficiently processed by combining an artificial intelligence algorithm, trend analysis and prediction are automatically carried out, and a reliable prediction result is generated.

The invention discloses an anomaly detection model selection method and system based on index portraits, and relates to the technical field of intelligent operation and maintenance of power systems. The method comprises the following steps: collecting historical data of a target monitoring index, extracting multi-dimensional features to construct an index portrait, and classifying the index portrait; screening candidate anomaly detection models from the matching rule base, performing adaptation degree scoring in combination with a model compatibility evaluation mechanism, determining an optimal anomaly detection model to perform anomaly detection, and outputting an anomaly judgment result; when a plurality of models exist, generating a final abnormal result through a confidence-driven arbitration mechanism; for multi-index abnormity, causal reasoning is carried out in combination with an electric power knowledge graph, main alarm indexes are determined, and secondary indexes are processed according to a delay strategy; meanwhile, incremental updating of index portrait features, adaptive adjustment of model parameters and dynamic optimization of matching rules are supported, and a whole-process closed-loop mechanism covering'portrait construction-model matching-result fusion-alarm decision-feedback updating 'is constructed.

The invention discloses a multi-temporal-spatial-scale cyanobacterial bloom early warning method for middle and large lake and reservoir water areas, and belongs to the field of cyanobacterial bloom prediction and risk monitoring. The method comprises the following steps: generating a pixel-level FAI index based on target water area remote sensing data, resampling meteorological data into a pixel level, then constructing a spatial-temporal distribution data set, training an Autoformer-ST-GNN time sequence model to realize FAI index prediction, dividing cyanobacterial bloom levels according to the FAI index prediction, and obtaining a global change trend; water quality monitoring points are arranged in key areas to collect data, historical water quality and meteorological data are utilized to train a DMC-PatchTST model fused with a blue-green algae migration period, and multi-time-scale prediction of the density of blue-green algae at the monitoring points is achieved; and finally, combining the global trend with a monitoring point prediction result to construct a space-time multi-scale cyanobacterial bloom comprehensive early warning system. According to the method, multi-source data and multiple models are fused, so that cyanobacterial bloom time-space multi-scale comprehensive early warning is realized, and the method is accurate and comprehensive.

The invention discloses a method for supporting fast switching of multiple models by NPU (Network Processing Unit) firmware. According to the invention, the system real-time performance and the resource utilization rate in a multi-task scene are obviously improved. Through a dynamic hierarchical caching strategy and a priority scheduling mechanism, the system can intelligently allocate cache resources and preferentially guarantee rapid loading of a high-frequency and high-urgency model, for example, in an automatic driving scene, switching delay of a path planning model can be reduced to a millisecond level, and non-perceptual switching of key tasks is guaranteed. The incremental parameter loading technology and firmware-level context management are deeply fused, only model difference data are transmitted, and hardware is utilized to accelerate and recover a calculation state, so that bandwidth waste caused by traditional full-amount loading is greatly reduced, the model switching efficiency of edge computing equipment is improved by more than 10 times when multiple tasks such as voice recognition and image processing are carried out in parallel, and the efficiency of the edge computing equipment is improved. And meanwhile, the calculation precision is kept lossless.

The invention relates to the technical field of computer science, in particular to a stamping prediction method based on a multi-machine learning model and an automatic compensation device.The stamping prediction method comprises the steps that firstly, a sheet microstructure and material characteristics are obtained through a transmission electron microscope and an atomic probe tomography technology, and then finite element simulation sampling is conducted; a data set is expanded by using an adaptive algorithm, multiple models are constructed and trained to obtain a final springback prediction model, an automatic compensation device is embedded to realize automatic compensation, and the automatic compensation device covers visual parameter input, embedded springback prediction, springback compensation calculation and an automatic control module. Parameters can be visually input, springback can be accurately predicted, a compensation value can be calculated, and the process can be adjusted. The method aims at solving the problems that in the high-strength plate stamping process, due to springback, the size and shape of a part deviate from the design requirement, and a traditional prediction and compensation technology is insufficient in precision and efficiency.

The invention relates to the crossing field of material design and artificial intelligence, in particular to a solid electrolyte intelligent inverse design method fusing a graph neural network and confidence analysis. According to the method, a prediction framework integrating multiple models is constructed, support vector regression, gradient boosting regression, a deep neural network and a graph neural network are included, component, process and structure parameter characteristics are fully fused, and the nonlinear mapping relation between input variables and performance parameters such as resistivity and conductivity is efficiently learned. In order to improve the credibility, a Bayesian neural network and a Monte Carlo method are further introduced, a confidence interval corresponding to each group of prediction results is output, and quantitative evaluation of the credibility of the prediction value is realized. In the inverse design module, high-dimensional submerged space parameters are generated based on a variational auto-encoder, and intelligent recommendation of parameter combination driven by target performance is realized in combination with strategies such as Bayesian optimization and a genetic algorithm. The design efficiency of the solid electrolyte and the success rate of material discovery can be effectively improved.

The invention provides a marine ranch water quality parameter multi-model fusion dynamic prediction method and system, and relates to the technical field of water quality prediction, and the method comprises the steps: obtaining data through multiple monitoring platforms, building a combined prediction model through LSTM, PSO-SVM and a random forest regression model, and dynamically adjusting the combined weight according to the prediction error of each model, and accurate prediction of the dissolved oxygen concentration is realized. Compared with single model prediction, the advantages of multiple models are fused, and the prediction precision is improved; the adaptability and robustness of the model are enhanced by a strategy of dynamically adjusting the weight; according to the method, the time sequence analysis and the early warning threshold are combined, the early warning of the oxygen deficit risk is realized, the position and range of the oxygen deficit risk area can be determined according to the spatial distribution characteristics of the prediction result, the accurate oxygenation of the oxygenation equipment is guided, the water quality safety of the marine ranch is effectively guaranteed, and the economic loss caused by oxygen deficit is avoided.

The invention discloses an interactive simulation method for calculating aging parameters of a superconducting cable and a superconducting cable system.The system comprises the steps that a cable body electromagnetic transient model, a cooling system transient thermal model and a three-dimensional multi-field finite element model are built, and cable material characteristics, environment parameters and operation condition data are input; a model interaction interface is constructed through a simulation platform, data interaction and coupling among multiple models are realized, simulation parameters are dynamically adjusted, accurate matching of model step length and dynamic characteristics of the system is ensured, and aging parameters of the superconducting cable system under complex working conditions are calculated. According to the method, the operation state of the superconducting cable is evaluated by analyzing the mixed data, aging analysis of the superconducting cable system under multiple working conditions is carried out, and decision support is provided for system maintenance.

An embodiment includes a method of augmenting performance and compliance of language model-based copilots. The method includes receiving application-specific guidance providing instructions that restrict responses output by an application-specific copilot based on a large language model (LLM). The method includes communicating to the LLM the application-specific guidance and setting an initial set of model parameters for the LLM. The method includes sequentially optimizing model parameters related to multiple model output characteristics of the LLM to generate a final set of model parameters. The method includes communicating the final set of model parameters to the LLM such that the final set of model parameters is implemented in the LLM during operations implemented by the copilot. The method includes deploying the copilot in an environment such that the copilot receives an actual query and replies with an actual response based on the LLM implementing the final set of model parameters.

The invention discloses a monitoring and early warning method and system based on active load prediction of a power distribution network, which breaks through the limitation of purely depending on historical data and simple statistical analysis, fully considers the complexity and uncertainty of load change influenced by various factors, and achieves the early warning of the load change through the combination of multiple models. The method can accurately classify and analyze the operation active load of the power distribution network through an active power analysis model, a decision tree model and the like, thereby improving the prediction precision. In the aspect of data processing, mass operation data generated by intelligent equipment in the power distribution network can be effectively integrated and mined by utilizing a vector machine model and a cloud large model server, an early warning mechanism is flexible, the method can adapt to different operation scenes and load characteristics, false alarm and missing alarm are avoided, and the method is suitable for popularization and application. According to the method, the power distribution network fault identification efficiency of the power distribution network power factors in different time periods is remarkably improved, and effective operation of the power distribution network is ensured.

The invention discloses a multi-object and multi-model deployment and control early warning system. The multi-object multi-model deployment and control early warning system comprises a multi-source data fusion module, an efficient algorithm and model optimization module, a real-time data processing and analysis module, a multi-model cooperation and decision module, an adaptive learning and evolution module and a security and privacy protection module. And a more comprehensive and deep judgment result is provided, so that the police officers can make more accurate decisions in the operation process. The multi-model structure of the system realizes cooperative work through the support of each module, and finally achieves the goals of accurate data, timely analysis and reliable early warning. And due to the close relationship between each module and multiple models, the system has higher decision-making ability and adaptability when facing a complex dynamic environment.

The invention relates to the technical field of road maintenance intelligence, and discloses a road intelligent maintenance decision-making system based on multi-dimensional evaluation and deep reinforcement learning, and the system comprises a geographic space data preprocessing module, a fuzzy TOPSI S evaluation module, a context awareness DQN strategy module, a credibility driving recommendation module, and a security strategy library module. The method comprises the following steps: generating a road health degree and a clustering label through multi-source data geographical weighted preprocessing and fuzzy TOPSI S dynamic weight evaluation; a reinforcement learning reward function is configured based on label differentiation, and a strategy space is explored in combination with noise; the confidence is verified through multiple models, a historical security policy is matched, and model optimization is driven through priority sampling injection samples. According to the method, the evaluation precision is improved through entropy weight-clustering dynamic weight, and flexible decision is realized in combination with reinforcement learning; and three-level security verification and closed-loop optimization are constructed to ensure that the risk is controllable, historical experience migration and multi-source data expansion are supported, and the cross-scene adaptive capacity is enhanced.

There are provided a method for designing a spectacle lens and a related technique thereof, the method including: a modeling step of dividing a pattern into multiple models with a deviation from a normally worn state of the spectacle lens as decentering, in which a reference model, a decenter model, and a tilt model are prepared as the multiple model; a sensitivity calculation step of calculating a decenter sensitivity and a tilt sensitivity; and a design step of using a base curve value in the vicinity of a balance solution as a base curve of the spectacle lens, wherein a base curve value c [unit: diopter (D)] which is a curvature of a surface in a region H where the on-retinal non-convergence region is not provided on an object-side surface is taken as x-axis, and the decenter sensitivity and the tilt sensitivity [unit: diopter (D)] are taken as y-axis, and an intersection of a plot of the decenter sensitivity and a plot of the tilt sensitivity is taken as the balance solution.

The invention provides a real-time intelligent agent dynamic scheduling method and system composed of multiple large models, and is applied to the technical field of data processing. The method comprises the following steps: performing dynamic scheduling expansion around a plurality of large-model real-time intelligent agents, firstly receiving task request data of a real-time service scene, performing grouping according to task types, priorities and resource requirements, pre-configuring a scheduling strategy, and generating a pre-scheduling queue and resource pre-planning information; and in combination with the target configuration information, scheduling task information after secondary optimization is obtained through multi-dimensional optimization such as task splitting and priority rearrangement. Dynamically adjusting and generating a target scheduling instruction based on real-time system resources and a model operation state, and cooperatively processing the target scheduling instruction with the large model capability adaptation parameters to form scheduling execution information; and finally, issuing to model nodes according to preset service quality requirements and resource constraints, generating information such as adaptive parameters and priority rules through feature extraction, association processing and the like, and supporting multi-model cooperation to efficiently complete real-time service tasks.

The invention discloses a multi-unmanned aerial vehicle cooperative positioning method and application, and relates to the technical field of unmanned aerial vehicle control and navigation, and the method comprises the steps: constructing a three-dimensional visual simulation platform, including dynamic environment modeling, sensor data simulation and virtual-real combination verification; identifying and tracking a target through a multi-sensor data fusion and deep learning algorithm of the three-dimensional visual simulation platform; based on a multi-unmanned aerial vehicle cooperative positioning algorithm, realizing real-time estimation of the position and state of the target through geometric modeling and interactive multi-model unscented Kalman filtering; information fusion and collaborative decision among multiple unmanned aerial vehicles are realized through a wireless communication network; and optimizing the flight path, task allocation and target identification strategy of the unmanned aerial vehicle based on simulation feedback. The problems that a traditional unmanned aerial vehicle cluster is low in positioning precision and poor in target recognition and tracking capacity in a complex environment can be solved.

A method and apparatus for video coding are disclosed. According to the method for the decoder side, a first syntax, related to whether the current block is coded in a CCLM related mode, is parsed from a bitstream comprising the encoded data for the current block. If the first syntax indicates the current block being coded in the CCLM related mode, a second syntax is parsed from the bitstream, wherein the second syntax is related to whether a multiple model CCLM mode is used or whether one or more model parameters are explicitly signalled or implicitly derived. The model parameters for the second-colour block are determined if the first syntax indicates the current block being coded in a CCLM related mode. The encoded data associated with the second-colour block is then decoded using prediction data comprising the cross-colour predictor for the second-colour block.

A computer-implemented system and method operate an automatic and adaptive animal behavioral system. The system receives user configuration and feedback, and generates a training set based on online animal behavior patterns and received user feedback. The system trains one or multiple models based on the training set, if the training set is sufficiently large. It then validates the models, and provides system generated classifications and action types and levels for animal's behavior patterns, if the models' accuracy rate is above a threshold.

The invention provides a large model ecological online evolution learning method based on end-cloud collaboration and an application thereof, and the method is applied to the field of e-commerce, so that efficient training and deployment of an e-commerce model are realized. According to the method, an end-cloud collaborative architecture is adopted, a cloud end deploys a large language model, an end side uses a small and medium-sized language model, and efficient training and reasoning of large model ecology are realized through collaborative work of the cloud end and the end side. In the training process, the cloud model serves as a teacher model, a pseudo tag is generated to supplement a data set, and optimization training of the end side model is guided. Meanwhile, a multi-model cooperation self-training mechanism is introduced, multiple models play different roles and mutually discuss, training labels are autonomously generated, and the end side model is further optimized. And meanwhile, user feedback information is brought into cue words of the cloud model for situational learning, so that the model performance is continuously optimized. The method is applied to an e-commerce scene, the e-commerce model ecology is successfully constructed and deployed, and intelligent development of the field is realized.

The invention discloses a system and method for estimating the charging capacity of an energy storage lithium battery, and particularly relates to the technical field of energy storage battery management, and the method comprises the following steps: collecting dynamic parameters in the charging process of the energy storage lithium battery, and forming a dynamic data set; calculating an environmental disturbance influence index and a charging response consistency factor based on time sequence analysis and feature decoupling; constructing a two-dimensional working condition mapping matrix, identifying a current working condition area and generating a scene label; determining parameter weights of the plurality of estimation models by adopting a probabilistic reasoning mode; according to the scene state, selecting a single model output result or fusing a plurality of model output results for estimation; according to the method, the environment disturbance influence index and the charging response consistency factor are constructed, so that the complex working condition is accurately identified; scene labels are automatically generated based on two-dimensional working condition mapping and a clustering algorithm, and a plurality of estimation models are dynamically selected or fused in combination with model confidence, so that the accuracy, robustness and intelligent level of estimation are improved.

The invention provides a fine control style migration mask assisted character cartoon generation method and device, and the method comprises the steps: 1, mask generation: generating different masks for the face and background of an original character image I, and setting style intensities for different regions; step 2, face processing: using a VToonify model GvToonify, retaining face details and applying a specified animation style to carry out style migration on the face to generate a stylized face image Iface; step 3, background processing: using a diffusion model Gdiffusion to carry out stylized generation on a background region, and generating a style background Ibg; according to the method, figures and backgrounds in one picture are generated respectively, key points are generated through mark removal by means of the masks, so that various models can only exert skilled parts of themselves in one picture, splicing is carried out at last, and finally, the face and the background are fused into a complete image # imgabs0 #. Therefore, the overall composition of the picture is optimal in multiple aspects.

The invention discloses a method for dynamically calling multiple models for a deep learning algorithm and a service architecture. The method comprises the steps that a client sends a request to a scheduling service through a standard AP I interface; after the scheduling service receives the request, key information of the request is analyzed and extracted to generate a task queue, and a Server server is selected; the scheduling service acquires state data provided by the Agent service in real time, and performs task allocation according to the load condition of the Server; the scheduling service stores the selected task information into a database, and marks a'to-be-processed 'state in a task state table; after receiving the scheduling task, the Server calls a model management service to load a specified reasoning model; and S6, after the reasoning task is executed, the reasoning service pool informs a scheduling service of a result through a callback mechanism. The method has the advantages that the processing capacity and efficiency of the deep learning task are improved, the requirements of different tasks are met by dynamically calling multiple models, meanwhile, resource use is optimized, and the expandability of the system is guaranteed.

A vehicle-mounted GNSS positioning method based on multi-motion model interaction includes: establishing a position-constant velocity (PCV) model and a position-constant steering angular velocity (PCSAV) model for two attitudes of a carrier (i.e., linear motion and turning motion) respectively to obtain a state estimation vector and a state transition matrix of the carrier of the PCV model and the PCSAV model at a previous moment, introducing an interacting multiple model (INM), establishing a heuristic position-velocity filtering (HPV)-IMM model based on the IMM model to achieve an information filtering interaction between the PCV model and the PCSAV model, and obtaining a state estimation vector and an error covariance matrix of the carrier at a current moment, so as to obtain a position and velocity of the carrier at the current moment. The present disclosure solves the problem of low accuracy of a traditional single kinematic model in multi-motion attitude vehicle positioning.

The invention discloses a multi-model-based complex scene path planning generation method and system, and relates to the technical field of path planning, a static environment information set and a dynamic environment information set in a road are collected in advance, and multi-layer feature representation of the road is constructed based on the static environment information set and the dynamic environment information set; and constructing a path planning model based on the multi-layer feature representation of the road, and generating a planned path by using the optimization model based on the path planning model. And the path planning capability of the self-driving automobile in a complex scene can be obviously improved.

The invention relates to a multi-model time division multiplexing and parallel loading reasoning service method and system, and the method comprises the steps: receiving a user model reasoning request, and searching a corresponding model according to the user model reasoning request; deploying part of parameter layers of the model required by reasoning into a GPU (Graphics Processing Unit); inputting the user model reasoning request text into a parameter layer deployed by the GPU for calculation; when the partial parameter layers are calculated, loading the residual parameter layers of the model to the corresponding specified GPU; after reasoning calculation is completed, the parameter layer is unloaded and returned to the CPU, and a reasoning result is returned. A GPU device cluster is used for carrying out reasoning service on the models, cooperation among device clusters of multiple models is achieved, and the reasoning time delay of each model request is reduced. And under the condition of reserving and occupying fewer system video memories, the same and even better reasoning speed can be achieved. The model deployment and scheduling process is automated, and the deployment cost of the model is reduced as much as possible under the condition that the SLO requirement of each model request is met.

The invention discloses a Java-based large language model dynamic adaptation method and system, belongs to the technical field of artificial intelligence, and aims to solve the technical problems of overcoming the interface difference of multiple model interfaces, simplifying the development process and providing standardized and high-expansibility large model management. Comprising the following steps: providing a model registration and dynamic loading service, a protocol conversion and parameter standardization service and a load balancing and failover service; a streaming transmission protocol service, a function call dynamic injection service and a global error processing mechanism are provided; when a user provides a document analysis and partitioning service, a vectorization index construction service and an RAG enhanced generation service to ask questions, relevant document blocks in the Pinecone are retrieved through the RAG enhanced generation service to serve as contexts to be injected into cue words of the large language model, and answers generated by the large language model are returned.