69 results about "Network agent" patented technology

The invention relates to the technical field of digital twinning, physical modeling and multi-source data fusion, and provides a digital twinning dynamic construction method based on multi-source data fusion and physical simulation. The method comprises the following steps: acquiring a multi-source heterogeneous data stream from a preset sensor array, a numerical simulation result and a historical database, identifying a key feature mode of a dominant physical process in the multi-source heterogeneous data stream, acquiring a key feature mode time-varying physical field evolution rule corresponding to the key feature mode by using a time sliding window and a forgetting mechanism, extracting a low-dimensional sparse characteristic parameter set reflecting dynamic behaviors from a high-dimensional observation space, and constructing a reduced-order proxy model by adopting Gaussian process regression, a neural network proxy model or an intrinsic orthogonal decomposition combined interpolation technology; and receiving a corresponding real-time observation data stream to establish a full-closed-loop feedback link from model prediction, high-fidelity solution verification to observation data correction in combination with the reduced-order proxy model so as to complete the construction of the digital twin.

The invention discloses a post-earthquake railway structure state evaluation method and system based on multi-source data fusion. The method comprises the steps that a finite element model is constructed according to the structure of a regional railway network, a simulation data set is generated, a physical information neural network model is trained, and a proxy model library is formed; acquiring multi-source observation data, and performing inversion according to the Bayesian theory, the structural damage parameters of the physical information neural network agent model and the multi-source observation data to obtain complete posterior probability distribution of the damage parameters; according to the method, statistical characteristics of damage parameters are extracted from posterior probability distribution, probability grading is carried out on the damage degree of the structure, driving constraint suggestions are generated according to the combination of damage probability grading and specifications, and Bayesian inversion time is reduced from several days to several hours through a physical information neural network agent model, so that the evaluation efficiency is improved; the output of the Bayesian method is probability distribution, the uncertainty range of the evaluation result is clearly displayed, and the uncertainty is quantified.

The invention belongs to the technical field of computational mechanics and artificial intelligence crossing in civil engineering, and discloses a foundation pit retaining wall deformation prediction method based on a physical information neural network, and the method comprises the following steps: 1, building a retaining wall physical information neural network agent model based on multi-source data fusion; 2, building a displacement-force inversion model of the retaining wall; and 3, carrying out iterative coupling solution on the support system. According to the method, mixed training is carried out by fusing physical rules and field measured data, so that the physical information neural network agent model follows a basic mechanics principle and can also fit specific engineering practice, the prediction result is more reliable, and compared with traditional finite element analysis, the efficiency is improved by several orders of magnitude; therefore, parameter optimization and real-time safety evaluation based on a large amount of calculation become possible.

The invention provides a pipe network model parameter inversion method and device based on an FMU standard and an optimization algorithm, and relates to the technical field of fluid simulation and parameter calibration. A pipe network physical model is established, and a to-be-inverted parameter is defined as an adjustable parameter and compiled and exported as an FMU format file; an FMU file is loaded in an optimization environment, a parameter sample set is generated through uniform random sampling, and a neural network agent model is trained to establish a mapping relation from parameters to system response; and a weighted error loss function is designed, deviation of measured data and simulation output is taken as an optimization target, optimal parameters of the neural network agent model are solved by adopting an optimization algorithm, and the pipe network system is operated according to the optimal parameters. According to the method, the combination of the Modelica modeling advantage and the Python optimization algorithm is realized through the FMU standard; optimization iteration is accelerated through a neural network agent model, and the parameter calibration efficiency is greatly improved; and high-precision automatic inversion of the key parameters of the pipe network is realized.

The invention provides a distributed scheduling method for railway traction stations and peripheral micro-energy grids, and belongs to the technical field of energy supply of micro-energy grid groups containing the railway traction stations, and the method comprises the steps: building a micro-energy grid sharing mutual aid grid structure containing the railway traction stations, obtaining a power model of each railway traction station, and optimally configuring energy storage equipment, obtaining operation constraints of the energy storage equipment; constructing an energy sharing benefit function by taking each micro-energy network agent as a game main body and taking individual benefit maximization as a target, setting constraint conditions, and establishing an energy sharing non-cooperative game model; and solving the energy sharing non-cooperative game model by adopting an alternating direction multiplier algorithm to obtain a distributed energy transaction behavior result and a price result. According to the method, distributed solution is carried out on the micro-energy network energy sharing non-cooperative game model based on the alternating direction multiplier method, original centralized optimization is converted into coupling optimization among a plurality of micro-energy network agents through interactive variables, the efficiency of the whole iteration process is guaranteed, and the privacy of each energy network is protected.

This disclosure provides systems, methods and apparatus, including computer programs encoded on computer storage media, for centralized management cloud connecting multiple enterprise networks. A network agent may be deployed within a private cellular network and act as an interface between the node(s) of the private cellular network and a cloud network controller. The network agent may obtain a local-based request to initiate a cloud-based procedure associated with network parameter(s), the network parameter(s) being associated with cloud network credential(s) that correspond to the private cellular network. The network agent may output a cloud-based request to the cloud network controller to initiate the cloud-based procedure. The cloud-based request may indicate at least a portion of the network parameter(s) and omit at least a portion of local credential(s) of the private cellular network. The cloud-based request may hide the cloud network credentials of the private cellular network from the cloud network controller.

The invention provides a suspension bridge cable clamp digital twin system based on wireless intelligent washers, and belongs to the crossing field of bridge engineering and intelligent monitoring technologies. According to the system, a wireless intelligent gasket is used as a core sensing unit, finite element analysis, a neural network agent model and a three-dimensional visualization technology are combined, and real-time mapping of a cable clamp physical entity and a virtual model is constructed; the system obtains data such as wireless intelligent gasket pressure, cable clamp-main cable friction force, cable clamp displacement and cable clamp structure node stress and strain through finite element analysis under different working conditions, and a parameter mapping proxy model is formed by training a mapping relation between the gasket pressure and other parameters through a neural network. And relevant models and data are developed and integrated by means of an Unreal Engine engine, and visualized real-time rendering is carried out, so that bolt fastening control in a construction stage and / or structural health monitoring in an operation and maintenance stage are / is realized. According to the system, digital management and control of the full life cycle of the cable clamp can be realized, and the monitoring precision and the engineering economy are remarkably improved.

The invention belongs to the technical field of wireless communication, and discloses an Internet of Vehicles optimization method and system based on edge computing and time division multiple access, and the method comprises the steps: carrying out the joint optimization of service deployment and replacement, RSU deployment, access control, and task unloading rates in VTUs and RSUs, and achieving the optimization of the Internet of Vehicles. And efficient, low-delay and stable task processing service is provided for vehicle-mounted communication in urban traffic, expressways or disaster areas. According to the method, a double-time-scale optimization strategy is adopted, the decision of service deployment and replacement is carried out on a coarse-grained time scale (frame), and other optimization problems are processed in a fine-grained time scale (time slot). The invention provides an innovative hierarchical DRL algorithm, and in the algorithm, a high-level agent adopts a deep Q network (DQN) proxy and is responsible for optimizing a service deployment and replacement strategy in a frame; and the bottom layer intelligent agent adopts a depth deterministic policy gradient (DDPG) for proxy, and specially solves related optimization problems of task unloading rate, access control and the like in the time slot.

The invention relates to the technical field of resource cache management, and discloses a page cache-based page loading method and device, electronic equipment and system.The method comprises the steps that after it is detected that a network agent tool meets a preset operation condition, a network request is intercepted through a capture event monitor of the network agent tool, and the network request is loaded to the electronic equipment; if the network request does not meet the special processing condition, judging whether the network request contains a target resource request, and if so, executing a resource response operation according to a resource loading strategy corresponding to the target resource request contained in the network request to obtain a first resource; otherwise, loading a second resource required by the network request through the network based on a preset network loading instruction; and displaying the target resource obtained by loading on the current page. Therefore, by implementing the method and the device, different resource loading and caching strategies can be adopted for different types of page resources, adaptive page caching is realized, the loading efficiency and stability of the page resources are improved, and the page browsing experience is improved.

The present application relates to a kind of shield tunnel irregular load inversion method based on neural network agent model, belong to shield tunnel load inversion field.The method, discard the assumption of load distribution mode in traditional lining load inversion method, inversion result can reveal the irregular load distribution mode of on-site tunnel, provide technical support for the service safety of tunnel service;Inversion is based on the deformation of lining whole circumference obtained by three-dimensional laser scanning technology, significantly increase the amount of field data, with the increase of n, the dimension of deformation vector can be increased sufficiently, avoid the defect that traditional inversion method easily leads to the non-uniqueness of inversion result.

The application discloses a power distribution area collaborative planning method and system based on a neural network agent model, and the method comprises the following steps: constructing a double-layer optimization model of flexible interconnection and energy storage collaborative planning of a power distribution area; randomly generating a candidate site and capacity scheme satisfying the constraint conditions of the double-layer optimization model, and performing optimization calculation of a running model on each candidate site and capacity scheme to obtain a real total cost of the whole life cycle corresponding to each candidate scheme, thereby forming a training sample set; performing supervised training on a preset neural network agent model; generating and iteratively updating a site and capacity scheme, evaluating the predicted total cost of the site and capacity scheme by using the neural network agent model, optimizing the site and capacity scheme according to the predicted total cost, and outputting a final target site and capacity scheme and a corresponding collaborative running strategy until a convergence condition is satisfied. The method effectively solves the problems of disconnection of 'planning-construction-operation', low calculation efficiency and the like in traditional planning, and realizes collaborative improvement of safety and solution efficiency.

The invention provides an agent model construction method, an automobile thermal management method and a related device, and the agent model construction method constructs a digital model of a to-be-simulated device, and carries out the calibration of the digital model through the pre-collected experimental data, and obtains a target digital model of the to-be-simulated device; acquiring simulation operation data of the to-be-simulated device by using the target digital model of the to-be-simulated device, and gathering the simulation operation data into a training data set; and training a pre-constructed neural network model by using the training data set to obtain a proxy model corresponding to the to-be-simulated device. By adopting the technical scheme of the invention, the neural network agent model of parts or systems related to automobile thermal management can be constructed, the simulation calculation load is reduced, and quick response is realized, so that the simulation efficiency of thermal management is improved, the development efficiency of a thermal management system is improved, and the efficiency of automobile thermal management is further improved.

The invention relates to the technical field of communication, and provides a group communication-oriented fixed-mobile network intercommunication method, device and equipment. The method comprises the following steps: a fixed network node establishes a first session with an intercommunication user plane node through a network agent module independently serving different groups, and an IP address field used by the agent module allows cross-group overlapping; the control plane node completes first session establishment and updates a network connection identifier replacement strategy; when the mobile terminals in the same group access the fixed network, the target IP is detected to fall into the overlapped address field, and a second session taking the intercommunication user plane node as an anchor point is triggered to be established; and the intercommunication user plane node establishes data forwarding association of the second session and the first session according to the matching relationship between the group and the IP, so that the mobile terminal directly accesses the fixed network terminal through a path formed by the third session, the second session and the first session initiated by the mobile terminal. According to the invention, through a centralized proxy, address multiplexing and session association mechanism, fixed-mobile intercommunication in an operator network is realized, and the resource efficiency and deployment flexibility are improved.

The invention discloses a multi-modal data driven complex product installation and adjustment feature and key event cognition method and system. The method comprises the following steps: firstly, synchronously collecting and performing time sequence alignment on vision, point cloud, force sense and motion data; extracting geometric invariant features decoupled from the visual angle through a dynamic graph convolutional neural network, and generating mechanical residual time sequence features based on a physical information neural network agent model; then, improving the heterogeneous features into structured scene semantic representation by using a multi-modal fusion network and an installation and adjustment domain knowledge graph; and finally, through a vertical domain large language model integrated with a temperature scaling probability calibration mechanism, real-time identification and quantitative confidence evaluation of installation and adjustment events such as stable slippage and abnormal clamping stagnation are realized. According to the method, the robustness of state perception, the physical interpretability of event identification and the credibility of decision making in the installation and adjustment process are effectively improved.

The invention discloses a hull motion digital twinning modeling method and system based on a time sequence circulation framework physical information neural network, and the method comprises the steps: building a neural network agent model which employs a framework combining a long-short-term memory network with a physical information neural network; comprising a time sequence feature extraction branch and a spatial feature coding branch, physical constraints are introduced, displacement of large-scale nodes does not need to serve as samples, and only a small amount of samples are needed to guide a model to learn a mechanical law; through the time sequence feature extraction branch, the high-precision and high-resolution dynamic deformation cloud picture of the whole structure can be generated only by relying on the input of a very few sensors. The invention provides a set of complete and landing technical scheme for realizing a low-cost and high-efficiency hull motion digital twin system.

Devices, systems, and techniques for mapping tokens generated by neural networks to application programming interfaces (APIs). In at least one embodiment, a processor includes one or more circuits that cause one or more tables to map one or more tokens generated by one or more neural networks to one or more tokens that are to be used by one or more APIs.

The application discloses a hull motion digital twin modeling method and system based on a timing cycle framework physical information neural network, wherein a neural network agent model is built, the neural network agent model adopts a framework of a long short-term memory network combined with a physical information neural network, includes a time sequence feature extraction branch and a space feature coding branch, by introducing physical constraints, without taking large-scale node displacement as samples, only a small amount of sampling is needed to guide the model to learn the mechanical law, and through the time sequence feature extraction branch, high-precision and high-resolution dynamic deformation cloud maps of the whole structure can be generated only by relying on the input of a small number of sensors. The application provides a complete and applicable technical scheme for realizing a low-cost and high-efficiency hull motion digital twin system.

A system performs operations of a neural network agent and a circuit simulator for simulating circuit sizing. The system receives input indicative of a specification and design parameters of a simulated circuit. The system iteratively searches a design space until a circuit size is found that satisfies the specification and the design parameters. In each iteration, the neural network agent computes measurement estimates for random samples generated in a trust region, which is a portion of the design space. Based on the measurement estimates, the system identifies a candidate size corresponding to an optimized value indicator. The circuit simulator receives the candidate size and generates a simulation measurement value. The system computes updates to weights of the neural network agent and the trust region based at least in part on the simulation measurement value for a next iteration.

The invention discloses a neural network proxy model-based power distribution area collaborative planning method and system. The method comprises the following steps of: constructing a power distribution area flexible interconnection and energy storage collaborative planning double-layer optimization model; candidate addressing and sizing schemes meeting constraint conditions of the double-layer optimization model are randomly generated, optimization calculation of the operation model is executed on each candidate addressing and sizing scheme, the real total life cycle cost corresponding to each candidate scheme is obtained, and a training sample set is formed; performing supervised training on a preset neural network agent model; and generating and iteratively updating the addressing and sizing scheme, evaluating the predicted total cost of the addressing and sizing scheme by using the neural network agent model, optimizing the addressing and sizing scheme according to the predicted total cost until a convergence condition is met, and outputting a final target addressing and sizing scheme and a corresponding cooperative operation strategy. The problems of'regulation-building-transportation 'disjunction and low calculation efficiency in traditional planning are effectively solved, and collaborative improvement of safety and solving efficiency is achieved.

The invention discloses a Bayesian polynomial chaos neural network proxy model method, and belongs to the technical field of sandwich board structure optimization design and proxy models. The method comprises the following steps: firstly, determining structural composition, size association and design parameters of the Y-shaped sandwich panel, constructing an automatic modeling script file, and simulating a drop hammer impact experiment process; secondly, generating a data set required by optimization of the Y-shaped sandwich panel, completing preprocessing, determining a protection performance evaluation index and an optimization target, and generating effective sample data; thirdly, training the proxy model based on the training set and the verification set; finally, multi-objective optimization design is carried out, design space is explored, a Pareto solution set is obtained, and a comprehensive optimal solution is selected. Through cooperation of the high-precision efficient proxy model and the intelligent optimization algorithm, the design period is shortened, the design space exploration range is expanded, the energy absorption protection effect is remarkably improved while the light weight of the Y-shaped sandwich panel is achieved, and reference is provided for engineering application and optimization design of the Y-shaped sandwich panel.

This invention relates to the field of bridge structural health monitoring systems based on the Industrial Internet, specifically disclosing a bridge structural health monitoring system based on the Industrial Internet. The system includes a bridge structural sensor network, a micro-meteorological sensing unit, a computational fluid dynamics simulation engine, a physical information neural network proxy model, a real-time boundary correction unit, and a structural health assessment center. By fusing micro-meteorological data with high-fidelity wind field simulation, a wind pressure-structural response mapping model with physical consistency is constructed, and real-time boundary correction is used to achieve closed-loop synchronization between the digital twin and the physical entity. The structural health assessment center identifies anomalies based on the expected response and measured deviations, and reduces false alarms by combining an adaptive threshold mechanism. Furthermore, the model has the ability to extrapolate and predict wind-induced vibrations, enabling early risk warnings. This invention can achieve high-precision damage identification and proactive safety warnings, improving the proactive safety assurance and intelligent operation and maintenance level of bridges in complex wind environments.

This application relates to the field of microgrid technology and discloses a method and system for setting droop control coefficients based on a physical information neural network. The method first obtains the target steady-state output current to be set for each parallel converter in a multi-source parallel DC microgrid; then, it inputs the target steady-state output current into a pre-trained physical information neural network surrogate model. This surrogate model is an inverse mapping model that uses the converter's steady-state output current as input features and the droop control coefficient as the output target. Its training loss function includes a physical constraint loss term constructed based on Kirchhoff's voltage law for the DC microgrid topology; finally, it outputs the corresponding droop control coefficients through model forward propagation, thus completing the parameter setting. This application improves the accuracy and efficiency of droop control coefficient setting, ensuring the accuracy of current distribution and operational stability of the DC microgrid.

The invention provides a complex business arrangement method based on a service grid, which comprises the following steps that: a plurality of business services are deployed on a container management platform in a containerized application form, the service grid is operated on the platform, and the business services are associated with network agents; a declarative business process definition is obtained by an arrangement control module, and the definition describes an execution dependency relationship among business task nodes; the arrangement control module translates the business process definition into a group of network flow management rules and a group of declarative resource definitions; and issuing the network flow management rule to a control plane of the service grid to drive flow circulation, and issuing the declarative resource definition to a container management platform to execute a specific service task. According to the method and the system, process execution is sunk to infrastructures, thorough decoupling of service logic, process logic and resource scheduling is realized, a technical foundation is laid for constructing an elastic and robust distributed service system, and the development efficiency, reliability and maintainability of a complex service system are remarkably improved.

The application discloses a post-earthquake railway structure state evaluation method and system based on multi-source data fusion, which comprises the following steps: constructing a finite element model according to the structure of a regional railway network, generating a simulation data set, training a physical information neural network model, and forming a proxy model library; collecting multi-source observation data, inverting the complete posterior probability distribution of the damage parameters according to the Bayesian theory, the structure damage parameters of the physical information neural network proxy model and the multi-source observation data; extracting the statistical characteristics of the damage parameters from the posterior probability distribution, probabilistically grading the damage degree of the structure, generating driving constraint suggestions according to the damage probability grading and combining the specifications, and reducing the Bayesian inversion time from several days to several hours through the physical information neural network proxy model, thereby improving the evaluation efficiency; the output of the Bayesian method is a probability distribution, which clearly shows the uncertainty range of the evaluation result and quantifies the uncertainty.

The invention relates to a lake reservoir water taking optimization method and system based on an LSTM model, and the method comprises the steps: collecting multi-source data including water quality, hydrology and a wind field in real time, carrying out the data preprocessing, inputting a long-short-term memory neural network agent model, generating time sequence feature information, inputting a pre-constructed actor reviewer model, and carrying out the time sequence feature information; an optimal water taking scheme is obtained; the actor commentator model comprises a value network and a strategy network; training the actor reviewer model in combination with a target network mechanism and an experience playback mechanism; and based on the optimal training parameter combination, training the actor reviewer model in a reinforcement learning environment composed of a state space, a reward function and an action space to obtain a trained actor reviewer model. Compared with the prior art, the optimization result and the mechanism model calculation result are high in consistency, the feasible optimal water taking scheme can be rapidly output, and the water supply requirements of lakes and reservoirs in the water taking opening water quality abnormal period are effectively met.

The invention belongs to the technical field of intelligent traffic, and particularly relates to a vehicle-road cloud integrated communication system and method based on environment backscattering. Comprising the following steps: step 1, constructing a vehicle-road-cloud integrated simulation environment; 2, defining reinforcement learning decision elements; step 3, cloud strategy training and model issuing; 4, vehicle online decision making and communication simulation are carried out; 5, strategy closed-loop updating and performance evaluation are carried out; according to the invention, an AmBC module is integrated at a vehicle end, and a lightweight deep Q network agent is deployed at a cloud platform, so that a closed-loop optimization mechanism of cloud training-road distribution-vehicle execution is realized; according to the invention, vehicles are allowed to carry out interference-free communication by reflecting RSU broadcast signals when cellular users occupy frequency spectrums; meanwhile, the DQN strategy dynamically selects three modes of'idle / active V2X / AmBC ', and the long-term average throughput is maximized.

The invention relates to transmitting time-critical data within an industrial automation system in which a plurality of redundant control applications (111-113) generating control commands to be executed by associated automation devices (101) is provided by means of software containers each running in a runtime environment or by means of a virtual machine. At least one network proxy (132) aggregates control commands from redundant control applications (111-113) into resulting control messages (1) and transmits the control messages to the automation devices (101) via a network (130) of the industrial automation system. In response to the control messages (1), the at least one network proxy (132) receives status messages (2) from the automation devices (101), multiplexes the status messages, and forwards duplicates thereof to the associated redundant control applications (111-113). For a plurality of possible locations of the at least one network proxy (132) within the network (130) and for a plurality of network usage scenarios, delays and / or path costs occurring when transmitting said control (1) and status messages (2) are simulated. One of the plurality of locations providing for optimal delays, paths costs and / or a combination thereof in the plurality of network usage scenarios is selected for placing the at least one network proxy (132). An execution of the control applications (111-113) is assigned to selected available hosts (121-123) in the network (130) depending on the delays, paths costs and / or the combination thereof.

The present invention belongs to the technical field of network management, and relates to a non-stress measurement method for the availability of network agent service (s), and system thereof. By adopting advanced data sampling, time series prediction and machine learning technology, the present invention can evaluate the availability of proxy services in real time and accurately, which is helpful for network administrators to better monitor, maintain and optimize proxy services and improve the overall security and stability of the network.

The invention provides an agent-oriented website map construction method. The method comprises the following steps: firstly, defining a design criterion of network agent website map service, then providing a website map building module based on the design criterion, and outputting the website map service which can be adopted by an LLM-driven agent; the website map building module is composed of three stages: a first stage: website structure analysis: finding a core page by traversing a hyperlink structure of a website, carrying out breadth-first search by using a crawler, controlling an exploration range by adopting a configurable scoring function, and eliminating noise through a dynamic content and navigation noise filtering algorithm; in the second stage, user exploration is simulated, Playwright is used as an automatic agent, an interactive element exploration and dynamic content triggering algorithm is designed to trigger loading of dynamic content and reveal hidden elements, and dynamic content capture or clean page snapshot is carried out; and a third stage: browsing the content annotation. The capability is enhanced by providing external knowledge, so that the computing resource and time cost of technology implementation is greatly reduced.