31 results about "Virtual intelligence" patented technology

The invention provides an intelligent ship digital twinning motion modeling and parallel anti-interference path tracking controller. The intelligent ship digital twinning motion modeling and parallel anti-interference path tracking controller comprises a guidance signal module, a speed optimization module, a parallel path tracking controller module, a data-driven learning predictor module and an actual controller module. The data-driven learning predictor module receives position information, course angle information and speed information from the actual intelligent ship system module, and the data-driven learning predictor module receives parallel path tracking control force and torque from the parallel path tracking controller module; the data driving learning predictor module sends estimation information of the position and course angle of a virtual intelligent ship and estimation information of the speed of the virtual intelligent ship to the guidance signal module and the parallel path tracking controller module. According to the method, the self-adaptive capacity of the ship under the complex sea condition is enhanced, the collision risk is reduced fundamentally, and an effective solution is provided for the navigation scene with the high safety requirement.

Aspects of the subject disclosure may include, for example, instantiating a virtual smartphone in a cloud infrastructure, installing a smartphone application on the virtual smartphone, receiving input sensor data from a physical user device, providing the input sensor data to the smartphone application on the virtual smartphone, receiving output data from the smartphone application on the virtual smartphone, and providing the output data to the physical user device. Other embodiments are disclosed.

The invention discloses an automatic driving test method and system. Belongs to the technical field of automatic driving testing. According to the method, a high-fidelity virtual environment is constructed by collecting real scene data, and a virtual sensor is configured; mapping the real track to a virtual space, and generating a dynamic traffic flow in combination with a car following and lane changing model; after the real vehicle and the virtual intelligent agent are synchronously initialized, the real vehicle runs in the physical field and periodically updates the state; the virtual traffic flow is dynamically evolved according to the state of the main vehicle, and the virtual sensor senses the target in the field of view and generates a virtual sensing result; real sensor data is used as a reference, and virtual and real perception information is fused to form unified input; and circularly executing until the task is completed, wherein each virtual agent only participates in one test. According to the scheme, high-reality and high-efficiency closed-loop testing is achieved, the defects of a traditional method in the aspects of scene richness and testing safety are overcome, and the reliability and repeatability of verification of the automatic driving system are improved.

The invention belongs to the technical field of virtual intelligence, and particularly relates to an intelligent manufacturing factory simulation teaching system based on AI + VR and an implementation method. The system comprises a VR display module, an AI learning guiding module, a factory simulation module, an interaction control module and a data storage module, all the modules achieve data interaction and function cooperation through standardized interfaces, the AI learning guiding module obtains an operation instruction from the interaction control module and sends a visual prompt instruction to the VR display module, and the VR display module displays the visual prompt instruction. Interacting trainee ability data with the data storage module; the factory simulation module receives an operation instruction of the interaction control module and provides equipment state data for the AI learning guiding module; the interaction control module sends a control instruction to the factory simulation module; the data storage module interacts with the AI learning guiding module through an API interface. According to the invention, the problems of lack of intelligence and adaptivity in teaching and lack of deep fusion between a virtual environment and a real industrial control protocol are solved.

The present disclosure provides an object identification method and device, equipment and storage medium, relates to the technical field of computers, in particular to the fields of big data, intelligent search and the like. The specific implementation scheme is: according to historical interaction data of a plurality of users, determining a candidate user in the plurality of users; according to the historical interaction data of the candidate user, determining an interaction parameter of the candidate user and a virtual intelligent user; according to the interaction parameter, obtaining interaction data of the virtual intelligent user and the candidate user; according to the interaction data, determining a category of the candidate user, the category of the candidate user indicating whether the candidate user has a preset behavior, and the object being the candidate user having the preset behavior. The identification of the object can be completed for the candidate user suspected of having the preset behavior without affecting other users, and the identification efficiency is also high.

The invention belongs to the technical field of virtual agents, and discloses a virtual agent driven hidden data collection method and system, and the method comprises the steps: constructing a virtual agent with environment perception and autonomous decision-making capabilities, and combining the association mapping of the virtual agent and an entity agent to achieve the intelligent driving of data collection; the method comprises the following steps: dynamically obtaining a burying point position and mode based on environment perception and target analysis, fusing multi-modal burying point data through a four-dimensional data model, and cooperatively collecting through a heterogeneous intelligent agent to form a unified data view, so as to enhance the integrity and consistency of data; according to the method, efficient, accurate and safe hidden data collection and integration are achieved, data values are fully mined, and the method is suitable for multi-scene data collection requirements.

The invention discloses a wind power plant optimal frequency track cooperative control method and system based on comparative clustering and hierarchical reinforcement learning. The method comprises the following steps: firstly, dynamically screening frequency-adjustable fans; a feature extraction network is pre-trained on massive historical data by adopting comparative learning, and the fan state is mapped to a low-dimensional feature space; performing online density clustering in the feature space by using an adaptive density clustering algorithm to automatically form a group in any shape, and enabling undistributed fans to serve as noise points not to participate in control; each group is used as a virtual agent, a theoretical optimal frequency trajectory is used as a tracking target, and an inter-group cooperation strategy is trained through a multi-agent depth deterministic strategy gradient algorithm; and the lower layer dynamically distributes power according to rotor kinetic energy and rotating speed margin. According to the invention, the cooperative tracking of the optimal frequency trajectory is realized, the lowest point of the system frequency is obviously improved, secondary falling is avoided, the structure is simple, the training is efficient, and the engineering practicability is good.

The invention relates to a multi-agent video multi-round question and answer method based on implicit knowledge enhancement, and the method comprises the steps: merging the video features of a real agent and the features of a virtual agent, and generating an extended agent feature set; constructing a multi-agent video question and answer model; training a multi-agent video question and answer model based on the extended agent feature set to obtain a trained multi-agent video question and answer model; and inputting the video features of the plurality of real agents and the to-be-processed question text features into the trained multi-agent video question and answer model to obtain a final video question and answer result. According to the method and the device, support features of historical rounds are expressed as virtual agent features, and unified modeling is carried out on the virtual agent features and real agent features of the current round, so that efficient multiplexing of context information in multi-round questions and answers is realized. The method has remarkable advantages in question and answer accuracy, calculation efficiency and communication overhead, and a new solution is provided for multi-round multi-modal question and answer tasks.

This invention discloses a multi-objective collaborative optimization control method for energy storage systems in microgrids, comprising: constructing a digital twin of the physical microgrid; defining a first virtual agent and a second virtual agent within the digital twin, wherein the first agent optimizes for long-term economic efficiency, and the second agent optimizes for short-term stability; based on real-time collected load data, renewable energy generation forecast curves, and equipment status, the two agents engage in game negotiation through a rolling time window mechanism until the negotiation process satisfies Nash equilibrium constraints; and sending the power command reached through the game to the power conversion unit of the physical energy storage system for execution via an encrypted communication protocol. This invention constructs a virtual model synchronized in real time with the physical microgrid and achieves precise and efficient control of the energy storage system through game negotiation between two virtual agents with different optimization objectives. This solves the problem of balancing economic efficiency and stability in microgrid operation.

This application relates to a self-optimizing control method and system for the operating state of an AI smart gateway. The method includes: preprocessing the real-time operating data of the gateway to obtain a unified state representation; calculating conflict features based on the covariance of historical multi-objective performance reward data; inputting the unified state representation into a target-specialized virtual agent to generate preference suggestions; extracting dynamic environmental features; inputting these and the conflict features into a meta-coordinator network to output jointly optimized actions; constructing an experience tuple; triggering a rapid adaptive mechanism when the environment changes abruptly; freezing the parameters of each target-specialized virtual agent; and updating the meta-coordinator network. This method, by constructing a hierarchical architecture of target-specialized virtual agents and a meta-coordinator, and introducing conflict feature perception and a meta-learning rapid adaptive mechanism, achieves dynamic modeling of multi-objective conflict relationships and rapid policy transfer under sudden scenarios, thereby improving the efficiency of multi-objective collaborative optimization and adaptive decision-making capabilities of the AI ​​smart gateway in dynamically changing environments.

The invention discloses a hanging basket construction whole process early warning system and method integrating digital twinning and the Internet of Things, and relates to the technical field of bridge engineering and intelligent construction. The method comprises the following steps: constructing a high-fidelity digital twinborn body of a hanging basket structure, calling a preset working condition simulation prediction model to carry out multi-path evolution deduction and risk assessment according to a preset or dynamically generated future construction working condition based on the current state of the digital twinborn body, generating graded early warning information and decision suggestions, and feeding back the graded early warning information and decision suggestions to a physical construction end, according to the working condition simulation and prediction model, a hanging basket structure is discretized into an intelligent cellular network, virtual agents with different objective functions are introduced, and the emergence behavior and risk of a system under the construction working condition are predicted by simulating self-adaptive evolution between cells and a dynamic game between the agents. The fundamental change of hanging basket construction early warning from static monitoring to dynamic deduction, from a single threshold value to a system game, and from post-event alarm to beforehand intervention is realized.

The invention discloses an agent communication cross-domain service quality guarantee method, equipment and a medium, belongs to the technical field of intelligent communication, and aims to solve the technical problem that the existing agent communication protocol uses HTTP / HTTPS in an application layer, and the existing packet filtering and mapping mechanism based on IP quintuple cannot be distinguished. In order to guarantee the service quality of an end-side agent and a virtual agent in a mobile communication network and provide differentiated QoS for different AI services in a 5G-A / 6G network, the adopted technical scheme is as follows: the network issues a cross-domain QoS mapping rule to the end-side agent in an agent session establishment process and issues an N4 mapping rule to a UPF; an inter-domain mapping relation between the 5QI and the DSCP, namely a 5G domain and an IP domain, is embodied; when the QoS flow is mapped in the uplink and downlink directions, if it is detected that an application layer uses an HTTP / HTTPS protocol and indicates an AI service, the DSCP is mapped to 5QI according to a cross-domain QoS mapping rule / N4 mapping rule, and QFI information corresponding to the 5QI is carried in an uplink SDAP packet header and a downlink GTP-U packet header for flow identification.

This invention relates to the field of artificial intelligence interaction technology, and discloses a method and apparatus for establishing intelligent agent friendships based on near-field touch. The method includes: when a first hardware carrier detects a near-field touch event, generating a first temporary token and broadcasting the first temporary token to a second hardware carrier; when the second hardware carrier detects a near-field touch event and receives the first temporary token, generating a second temporary token and a pairing response credential; performing multi-dimensional verification based on the first temporary token, first touch event data, second temporary token, pairing response credential, and second touch event data; when the multi-dimensional verification passes, establishing a friendship relationship between the first and second intelligent agents, sending the binding result to the first hardware carrier, second hardware carrier, client, and corresponding virtual intelligent agent, and updating the interaction state. This invention reduces mispairing and improves pairing success rate and interaction security through bidirectional triggering, one-time tokens, and multi-dimensional verification.

The invention relates to a transformer substation monitoring system based on a master station and substation service interaction mechanism. The transformer substation monitoring system comprises an automatic equipment remote operation and maintenance master station and an automatic equipment remote operation and maintenance substation. The automatic equipment remote operation and maintenance substation communicates with the automatic equipment remote operation and maintenance master station and communicates with intelligent electronic equipment of a transformer substation, the automatic equipment remote operation and maintenance substation comprises an equipment virtual module, and the equipment virtual module comprises a plurality of virtual intelligent electronic equipment. And each virtual intelligent electronic device corresponds to each intelligent electronic device in the substation and is used for storing monitoring information of the intelligent electronic device. According to the invention, a servitization main station and substation data communication mode is adopted, and on-demand retrieval of various data of the substation by various services of the main station is realized.

The application discloses an AI driving-based video content automatic auditing method and system, relates to the technical field of video content analysis, and comprises the following steps: acquiring video information to be audited, and performing multi-level segmentation on the video information; in the slice, node simulation is performed according to the content independence of the video information; the analysis of associated branches and historical characteristics is performed on each node to generate attribute characteristics of the node; content auditing is performed through the auditing strategy of multiple virtual intelligent agents, and the analysis of a violation space is performed according to unstable parts in an auditing result; and a final video content auditing report is generated according to the auditing result and the analysis result of the violation space. The application improves the structured identification and intelligent auditing efficiency of multi-modal video content, realizes accurate positioning, label attribution and automatic early warning of complex violation behaviors, and enhances the content security management capability.

The invention discloses an unmanned aerial vehicle formation adaptive fault-tolerant control and collision avoidance method based on a hierarchical structure, and the method comprises the steps: constructing a trajectory planning layer system based on an all-drive system and an artificial potential energy function method, and employing the projection of the relative speed between virtual intelligent agents on the relative distance and the relative distance as the input of a fuzzy logic system, the intensity of the collision avoidance potential field is determined; a six-degree-of-freedom dynamical model of the fixed-wing unmanned aerial vehicle is divided into a translation dynamical subsystem and a rotation dynamical subsystem, and a control scheme is designed by adopting an all-wheel-drive system method; a novel self-adaptive disturbance observer is designed and is used for compensating adverse effects caused by composite disturbance under the condition of unknown disturbance boundaries; a rotational dynamics subsystem is divided into two channels, and reference signals of the channels are obtained based on a nonlinear guidance law and a coordinated turning principle. According to the method, the robustness, fault tolerance and collision avoidance stability of the fixed-wing unmanned aerial vehicle formation are remarkably improved.

The application discloses a kind of multi-wheel mobile robot modularization formation collision avoidance control system and method, including online regulator module and trajectory tracking controller module.First, a plurality of virtual intelligent agents are constructed in online regulator module, and formation collision avoidance reference trajectory satisfying the requirements of wheeled mobile robot system is generated;Then, the trajectory tracking controller module of wheeled mobile robot is designed, so that it accurately tracks formation collision avoidance reference trajectory, to realize formation collision avoidance control.The system and method of the application can perform formation collision avoidance task without changing the controller of wheeled mobile robot, enhance the generality and convenience of design, is the expansion of prior art method, and is also a kind of prediction to possible application direction of future multi-wheel mobile robot system.

The application discloses a distributed architecture-based intelligent terminal data synchronization scheduling management system and method, and relates to the technical field of terminal data management; the application establishes a virtual mapping platform to map intelligent terminal ports and build a virtual intelligent terminal node network; a terminal state directory is constructed to record the state of the intelligent terminal; the real-time transmission data of each intelligent terminal is compared and evaluated to obtain the homogenization degree of the real-time transmission data of each intelligent terminal; the transmission data of each intelligent terminal is analyzed in terms of space-time demand characteristics to determine the real-time data transmission demand and complexity distribution of each intelligent terminal; the online duration of each intelligent terminal is called, and the online resource ratio priority of each intelligent terminal is planned in combination with the space-time characteristic demand of data transmission, and the terminal state directory is updated; the application guarantees the real-time demand of the intelligent terminal to adapt to a dynamic network environment and diversified terminal behaviors, and realizes differentiated scheduling of network resources.

The invention relates to a virtual-real fusion asymmetric mapping method based on dual unmanned aerial vehicle cooperative control, and the method comprises the steps: building a dual-vehicle asymmetric mapping model for a single-agent task scene in a virtual-real fusion scene; the pose information of the two real agents is synchronized according to the task scene limitation and the task time, and the two real agents autonomously and alternately receive the information of the virtual agent sensor and alternately complete the specified task. The invention relates to synchronization of a virtual world coordinate system and a real world coordinate system and displacement rotation transformation when two real agent local coordinate systems alternately map the same virtual agent coordinate system. And setting a dual-computer alternative mapping mechanism to realize real-time mapping of a large-scale operation space. According to the method, the scene space scale limitation of the virtual-real fusion experiment when the physical space is insufficient is made up, and distortion caused by simple speed and length scaling transformation in order to realize mapping of a small space to a large space is avoided.

The invention discloses a high-precision map generation system and method based on a large model, and relates to the field of high-precision map generation, and the system comprises a multi-agent simulation verification ring which comprises a high-fidelity simulation environment based on neural rendering and a group of virtual agents with differentiated driving strategies; the multi-agent simulation verification ring is used for receiving initially generated high-precision map data, driving the virtual agent to perform preset mileage driving including an edge use case in the simulation environment, collecting an interaction conflict event and a track smoothness index of the virtual agent and the environment, and generating a targeted optimization instruction; the system is used for solving the technical problems that in the prior art, an active safety verification mechanism is lacked in the high-precision map generation process, elastic supply cannot be achieved according to user requirements, hidden safety defects in the map are difficult to mine and correct, and the overall resource utility is not optimal due to isolated optimization of all links of the system.

This invention relates to the field of virtual-real fusion interaction technology, and discloses a method, apparatus, device, and medium for virtual intelligent agent interaction based on a physical carrier. The method includes: acquiring interaction event data from a physical hardware device; encoding the interaction events in the interaction event data to obtain a unique event identifier; validating the interaction behavior based on the unique event identifier; when the validation passes, uploading the unique event identifier and event proof to a server via a user terminal; upon receiving the unique event identifier and event proof, the server performs a legality verification; if the verification passes, querying the target virtual intelligent agent based on the unique event identifier and a preset binding mapping relationship; and updating the interaction state of the target virtual intelligent agent according to the event type corresponding to the unique event identifier. This invention acquires NFC touch and inertial interaction events through physical hardware, generates unique event identifiers, and drives the virtual intelligent agent to update its state, thereby improving interaction security and preventing interaction misjudgments.

The invention provides a ship power system operation training evaluation method and device and electronic equipment, and relates to the field of data processing. In the method, a ternary time sequence flow is generated by collecting multi-source signals and performing semantic alignment, and a surface track and a deep track are extracted to establish a multi-view immersion link; extracting a target sub-graph based on a heterogeneous node graph structure in combination with a standardized process library, and generating a causal candidate link set and a key node sensitivity spectrum; injecting disturbance through the virtual agent and recording a training deviation type; on this basis, a dimensionalized score set and a global score abstract are formed, and in combination with an evidence coverage rate label and a time window confidence label, the flow compliance, the strategy switching agility, the anti-interference stability and the communication definition are quantitatively evaluated. By implementing the technical scheme provided by the invention, the evaluation accuracy of the operation training of the ship power system of the trainee can be improved.

Embodiments of the present specification provide a virtual agent training method and apparatus, a computing device and a storage medium, the virtual agent training method comprising: obtaining sample object data, the sample object data comprising sample object state data, sample object action data and sample object reward data of a plurality of time steps; the sample object state data of the multiple time steps are input into an initial evaluation network in the virtual agent, an evaluation value corresponding to each time step is obtained, and the evaluation values are used for evaluating the value of the sample object state data of each time step; determining parameter adjustment information corresponding to each time step according to the sample object reward data and the evaluation value of each time step; and training the initial strategy network and the initial evaluation network according to the parameter adjustment information, the sample object reward data of each time step and the evaluation value until a preset training stop condition is met, and obtaining a target strategy network and a target evaluation network.

A system for managing and optimizing real world entities with machine learning algorithms are described including a server computer system configured to store and process input data, the server computer system comprising a memory and a processor; and wherein the memory of the server computer system stores a persistent virtual world system comprising virtual replicas of the real world entities, and wherein the server computer system is configured to generate explicit data sets representing functioning and behavior of the real world entities; train the machine learning algorithms with the explicit data sets to generate trained machine learning data sets; and apply the trained machine learning data sets in an artificial intelligence application to manage operation of the real world entities and generate real behavior data of the real world entities during the operation of the real world entities. Corresponding methods are also described.

The invention relates to an intelligent agent data generation method and device, electronic equipment and a storage medium. The method can comprise the following steps: acquiring multi-modal interaction data and agent task information at the current moment; performing semantic analysis and state understanding on the agent task information to generate an agent action sequence; based on the agent action sequence, driving the body-equipped agent to perform interaction processing with the external environment to obtain interaction feedback information; carrying out fusion processing on the multi-modal interaction data, the agent action sequence and the interaction feedback information, and constructing agent fusion information; generating virtual agent scene information according to the agent fusion information; and based on the agent fusion information and the virtual agent scene information, strategy model parameters of the agent with the body are updated, and agent update strategy data are generated, so that a strategy model forms closed loop iteration. According to the technical scheme provided by the invention, the consistency of the multi-modal data, the time sequence synchronism and the model training stability can be improved.

This invention discloses a VR-based wayfinding system interaction method and server. The method first acquires and parses a scene scan dataset of the target virtual reality scene. Next, it creates a pair of interactive virtual agents, including an environment-building agent and a narrative-eroding agent, and configures an environment memory model and a fictional anchor point generation strategy for each. Then, it drives the interactive virtual agents in an adversarial game, iterating alternately until a preset equilibrium condition is met, outputting a game-based wayfinding path and a set of target fictional anchor points. Furthermore, it assigns matching fictional anchor points to wayfinding nodes and generates activation commands to combine them into a narrative wayfinding path. Finally, it performs wayfinding interaction rendering in the target virtual reality scene, overlaying virtual information content according to the fictional anchor point activation commands to output a blended image to the user device, thus improving the intelligence and personalization of wayfinding interaction in virtual reality scenes.

The invention relates to a human-in-loop virtual-real fusion digital twinning system, and belongs to the technical field of digital twinning. According to the method, the entity sand table system is used for simulating the entity environment and the entity agent of the complex operation system where the person is located in the loop, virtual simulation is carried out on the complex operation system where the person is located in the loop through the virtual agent and virtual personnel of the twin deduction system, and twin simulation can also be carried out based on state information of the entity sand table system; or the two are combined to realize virtual-real fusion simulation. According to the virtual-real fusion digital twinning system, an entity agent-virtual human agent virtual-real collaborative operation mode is achieved, a human agent performs virtual-real collaborative operation with other agents in a simulation deduction platform in a virtual-real fusion mode, and credible simulation of a person in a complex operation system where a loop is located is guaranteed.

In embodiments of the present invention, a virtual phone is provided which includes a touch screen mounted on an enclosure that includes a processor, memory, wireless communication port, and a power port. The virtual phone further includes a software application executing on the processor to control image display on the touch screen and emulate features of a handheld device presented in the image in response to a user touching a presented feature. The virtual phone further includes a configuration of a user-specific handheld device that is derived from the user-specific handheld device is stored in a memory that is accessible to the processor; wherein the virtual phone presents a visual representation of the user-specific handheld device on the touch screen based on the configuration.

Techniques are described for a virtual smart network interface card to facilitate data transmission in an edge device providing cloud-computing operations. A private virtual network data plane hosted within an execution environment of the cloud-computing edge device can receive a request to create a virtual network interface. The request can include a networking command. The private virtual network data plane can send the networking command to a command proxy module. The command proxy module can then use the networking command to generate the virtual network interface within a first networking namespace of a host operating system executing on the cloud-computing edge device.

This invention proposes a quality control system and method for external beam radiotherapy based on a virtual intelligent medical platform. The system includes: a mixed reality device for acquiring spatial information and patient images within the treatment room, and overlaying a virtual image interface providing instructional information onto the real scene; and a server, communicatively connected to the mixed reality device, sending instructional information to the mixed reality device, receiving images of the implementation process acquired by the mixed reality device, determining whether the implementation process meets quality requirements, and sending the determination result back to the mixed reality device. This invention achieves intelligent, end-to-end quality control of radiotherapy positioning and treatment without altering existing workflows. Intelligent guidance, inspection, and recording are implemented throughout the radiotherapy positioning process, enabling paperless operation and full traceability; intelligent patient monitoring is achieved throughout the treatment process, avoiding oversights caused by human intervention.