155 results about "Management agent" patented technology

The invention discloses a heat supply system multi-working-condition scheduling control method based on a large model and multiple agents. The method comprises the steps that a heat supply system digital twinborn model is established; the method comprises the following steps: setting a scheduling control center management agent, predicting a thermal load demand by adopting a large language model and combining a thermal load prediction model, dividing a system into a plurality of operation conditions, and setting a plurality of condition sub-agents corresponding to the operation conditions; the dispatching control center management agent judges whether the working condition sub-agents under the corresponding operation working conditions are triggered to start working or not, and meanwhile heat load demand values, predicted through a large language model, of all time periods in the future and the heat production amount of the new energy heat supply unit are transmitted to the corresponding working condition sub-agents; and the working condition sub-agent adopts a large language model to establish a system optimization scheduling control knowledge graph based on the large language model, obtains a scheduling control strategy of each device on the source network load side of the heat supply system, and feeds back the scheduling control strategy to the digital twinborn model for guiding the actual operation management of the heat supply system.

The invention relates to the technical field of artificial intelligence, can be applied to business system platforms of financial science and technology, medical treatment and health and the like, and discloses a task processing method, device and equipment based on multi-agent cooperation and a medium. And constructing a sub-task dependency graph, sequentially scheduling and executing the sub-tasks, and performing parameter completion to obtain the completed sub-tasks. And tracking the execution progress and the historical record of the completion subtask through a preset progress agent, and generating an executable operation decision by using a preset decision agent. If the task execution does not reach the expected effect, feeding back difference information and a correction suggestion, adjusting an operation decision and generating an updating operation; if the task achieves the expected effect, task completion information is sent to the progress agent, and the state is updated to be task completion. According to the method, the perception accuracy is improved, a task dependence tracking and feedback correction mechanism is provided, and a cross-application automatic task is successfully realized.

The invention discloses a protocol analysis method and device based on multi-agent cooperation and a medium, and belongs to the technical field of industrial internet. The method comprises the following steps: constructing and associating a distributed agent cluster and a global protocol knowledge base; preprocessing the original Internet of Things protocol based on the analysis agent to generate a protocol analysis task packet; processing the protocol analysis task packet based on the scheduling agent so as to distribute the protocol analysis task packet to the target analysis agent; performing feature extraction and protocol analysis on the analysis task packet through an analysis agent, and outputting structured analysis data; if the unknown protocol format is recognized, feature extraction is carried out on the unknown protocol format and the unknown protocol format is sent to the management agent; and processing the unknown feature vector through the management agent to determine a new protocol analysis rule of an unknown protocol format, and synchronizing the new protocol analysis rule to the global protocol knowledge base. According to the method, the technical effect of efficient dynamic analysis adaptation of the industrial Internet of Things protocol is achieved.

The invention discloses a data governance agent system and an implementation method, and relates to the field of data processing, and the system comprises a data analysis module, a data storage module, a user response module, a quality monitoring and feedback module and a field large language model module. The data analysis module performs unified analysis modeling on structured and unstructured data; the data storage module adopts a graph database and a vector database to store knowledge data; the user response module supports natural language interaction and intelligent question and answer; the quality monitoring and feedback module optimizes data quality through closed-loop feedback; and the field large language model module constructs an optimization model adaptive to the vertical field. Through multi-modal analysis, knowledge graph construction, double-library collaborative storage and model dynamic optimization, a data barrier is broken through, manual dependence is reduced, the data treatment efficiency, quality and intelligent level are improved, and the method is suitable for specialized data treatment requirements of industries such as manufacturing and finance.

Systems and methods for hierarchical workspace orchestration are described. In an illustrative, non-limiting embodiment, an Information Handling System (IHS) may include: a processor and a memory coupled to the processor, the memory having program instructions stored thereon that, upon execution by the processor, cause the IHS to: determine, by a local management agent configured to instantiate a primary workspace, that an operation involving a subordinate workspace instantiated by a peripheral device coupled to the IHS comprises sensitive data; and determine whether to process the sensitive data prior to execution of the operation.

Disclosed are various approaches for blockchain-enabled large language model (LLM) red teaming. A large language model (LLM) red teaming management agent executed in a distributed ledger environment can deploy a plurality of subsidiary agents in the distributed ledger environment. An LLM red teaming agent deployed in the distributed ledger environment can cause a network service to perform an LLM red teaming action and return an LLM red teaming result. The LLM red teaming result can be evaluated to generate endorsement status information for the LLM.

The invention provides a network fault repairing system and method based on large language model multi-agent collaboration, and the system comprises a multi-agent collaboration module which comprises a management agent and a plurality of execution agents which are in communication connection with the management agent, the management agent and each execution agent are respectively provided with a large language model; each execution agent comprises a detection agent, an analysis agent, a diagnosis agent, a processing agent and a verification agent; and the processing agent comprises a congestion processing agent and an interruption processing agent. According to the invention, intelligent and automatic closed-loop processing of the network fault from sensing to repairing can be realized through a multi-agent collaborative architecture.

The embodiment of the invention provides a conference management system and method fusing a large language model and multi-agent collaboration. The system comprises a multi-modal central language large model and a multi-agent platform, and the multi-modal central language large model comprises a multi-modal interaction module, a full-duplex interaction interface, a task reasoning module, a task planning module and an agent scheduling module; the multi-agent platform comprises a data searching and issuing agent, a agenda management agent, a voice transcription agent and a knowledge base updating agent. According to the embodiment of the invention, the conference process is divided into a plurality of independent intelligent agents, and then the intelligent agents are scheduled and dynamically arranged in a unified manner by the central big model. Strategies can be flexibly adjusted according to real-time situations, and intelligence and high adaptability of the whole process are guaranteed. The conference process and the emotional state are monitored. And the central big model sends an intervention instruction. The active conference management can be realized, a targeted intervention means is provided, and the conference efficiency and the communication effect can be improved.

Systems and methods are provided for facilitating the management of AI (artificial intelligence) agents, AI agent skills, and customer access to AI agents and AI agent skills. Systems identify an AI agent and AI agent skills that the AI agent can utilize when interacting with different customers. The systems also identify a plurality of customers that the AI agent is capable of interacting with while utilizing one or more of the AI agent skills. The systems also generate and display interfaces that identify the plurality of customers with interactive permission controls for selectively enabling and / or disabling AI agent interactions by the AI agent the corresponding customers.

The invention provides a Token compression method and device for large model dialogues, which can effectively reduce the number of Tokens (lexical elements) of texts input into a large model in multiple rounds of dialogues, and can also retain key information of early parts of historical dialogue texts at the same time. The method is applied to a Token compression management agent, and comprises the following steps: after obtaining a question text input by a user to a terminal, determining the number of lexical elements in an initial text (including a historical dialogue text and the question text) to be sent to a large model; if the initial text lexical element number does not meet the threshold requirement, switching the state into a lexical element compression state, interacting with the large model to obtain the lexical element number of compressed intermediate dialogue texts (parts except for the previous k rounds of dialogue texts, k is greater than or equal to 0), and keeping a semantically generated abstract text. And splicing the first k rounds of dialogue texts, abstract texts and question texts to obtain a target text. If the number of the target text lexical elements meets the threshold requirement, sending to the first large model.

This disclosure describes a large language model (LLM)-based agent operating system comprising an application layer, a kernel layer, and a hardware layer. The kernel layer comprises an AIOS kernel comprising: LLM core(s), an agent scheduler, a context manager, a memory manager, a storage manager, and a system call interface configured to manage interactions among the agent scheduler, the context manager, the memory manager, and the storage manager.

The invention relates to the technical field of cloud computing and artificial intelligence, in particular to an agent collaborative management system and method based on multi-tenant isolation and dynamic role permission, and the system comprises the following modules: a tenant authorization agent module, a role authorization agent module, a user personalized management agent module, and a cross-tenant collaborative security mechanism. A dynamic permission effective process; the method has the beneficial effects that the first-step distribution of the agents is realized through the tenant authorization agent module, the second-step distribution of the agents is realized through the role authorization agent module, and finally the capability of individually loading and using the agents by each user is realized through the user personalized management agent module.

Systems and methods for user steering through workspace orchestration are described. In an illustrative, non-limiting embodiment, an Information Handling System (IHS) may include a processor and a memory coupled to the processor, the memory having program instructions stored thereon that, upon execution by the processor, cause the IHS to: receive, at a local management agent from a workspace orchestration service, one or more files configured to enable the local management agent to instantiate a workspace based upon a workspace definition, where the workspace definition enables the workspace to communicate with a selected peripheral device.

Systems and methods for workspace orchestration based on energy scores are described. In an illustrative, non-limiting embodiment, an Information Handling System (IHS) may include: a processor; and a memory coupled to the processor, the memory having program instructions stored thereon that, upon execution by the processor, cause the IHS to: receive, at a local management agent from a workspace orchestration service, one or more files configured to enable the local management agent to instantiate a workspace, where the one or more files are selected or assembled based upon a workspace definition created, at least in part, based upon an energy score; and instantiate the workspace.

Systems and methods for mobile device management (MDM) agent rollback are disclosed. In some embodiments, a system configured to perform operations including: detecting a triggering event during operation of a first MDM agent; designating the first MDM agent as a bad agent; designating a second MDM agent as a new active agent; determining whether a first data generated by the first MDM agent is useable for continued operations by the second MDM agent; if the first data is determined to be useable by the second MDM agent, then activating the second MDM agent for continued operations using the first data; and if the first data is determined to be not useable by the second MDM agent, then activating the second MDM agent with known good data.

The invention belongs to the technical field of building energy intelligent management and control, and discloses a model-data dual-driven building energy management and control agent training environment creating method. The method comprises the following steps: firstly, constructing a thermodynamic mechanism model of a building envelope and a simulation model of a heating ventilation air conditioning system, and coupling the two models through a joint simulation interface; then, cleaning and repairing processing such as time sequence alignment, missing value filling and abnormal value filtering is carried out on the existing building operation data, and the data quality is ensured. And then, running the joint simulation model to generate output data such as hourly load and equipment state. And finally, through space-time label association and feature extraction, performing multi-source fusion on the cleaned and repaired measured data and model output data, constructing a training database containing multi-dimensional features such as building thermal dynamic response and equipment operation features, and providing high-precision and full-scene training data support for a building environmental control agent.

The invention discloses a learning and optimization-based power grid interaction type residential building online cooperative operation method, which belongs to the technical field of power grid interaction type residential building cooperative operation and artificial intelligence crossing, and comprises the following steps of: establishing a problem of minimizing the operation cost of a single residential building during a non-power grid service period; modeling a single residential building operation cost minimization problem as a Markov decision process; training a residential building energy management agent by adopting a cross normalization algorithm to obtain an energy management strategy; total power limitation set by a power distribution system operator is considered, and a resident building group double-layer collaborative operation cost minimization problem is established; solving the collaborative operation cost minimization problem based on the energy management strategy obtained by each residential building and optimization auxiliary binary search, and executing a decision; the power distribution system operator calculates economic compensation and issues the economic compensation according to the energy management decision of each residential building before and after collaboration, the energy cost can be effectively saved, and the higher power grid service capability is achieved.

Systems and methods for securely deploying a collective workspace across multiple local management agents are described. In some embodiments, an Information Handling System (IHS) may include a processor and a memory coupled to the processor, the memory having program instructions stored thereon that, upon execution, cause the IHS to: receive, at a workspace orchestration service from a first local management agent, first context information and a first split key; receive, at the workspace orchestration service from a second local management agent, second context information and a second split key; determining, by the workspace orchestration service, that the first and second context information match a collaborative workspace policy; in response to the determination, authenticate the first and second split keys; and in response to the authentication, transmit a collaborative workspace definition to the first and second local management agents.

The invention discloses an open pit coal mine ground production system collaborative operation method based on an Agent simulation technology, and belongs to the technical field of coal storage and transportation. The method comprises the following steps: constructing a data acquisition and environment perception Agent, an anomaly detection and coordination management Agent, an optimization calculation Agent and a control execution Agent; the method comprises the following steps: collecting coal flow, coal type characteristics and equipment operation data of a conveying system in real time, and constructing a multi-objective optimization function taking energy consumption minimization, space utilization rate maximization and equipment load balance as objectives in combination with bin space capacity, coal pile three-dimensional form and environmental parameters of a storage system; an improved particle swarm optimization algorithm is adopted to solve optimal control parameters, and the belt conveyor speed, the coal type distribution priority and the emergency path are dynamically adjusted; through a closed-loop feedback mechanism, material blockage, shipping space overrun and equipment faults are monitored in real time, a self-healing strategy is triggered, and the model is updated. The hysteresis control defect of a traditional storage and transportation system is overcome, the operation efficiency is remarkably improved, and energy waste is reduced.

Implementations for chiplet fault management through hierarchical error reporting and coordinated recovery in an accelerator interconnect. In some implementations, a system for a UALink-based network comprises a chiplet die coupled to an accelerator die, a management agent on the accelerator die, and a centralized controller. Upon fault detection, the chiplet die transmits a fault notification to the accelerator die, the management agent reports the fault to the centralized controller, and the centralized controller determines a recovery action based on the fault identifier and severity. In some implementations, the accelerator die reports the fault through dual paths, to a host processor via CXL or PCIe error reporting and to the centralized controller via the management agent, and the host processor and the centralized controller coordinate recovery, with the host managing workload-level recovery and the centralized controller managing network-level recovery.

Systems and methods for providing trusted local orchestration of workspaces are described. In some embodiments, an information handling system (IHS) can include: a processor; and a system memory coupled to the processor, the system memory having program instructions stored thereon that, when executed, cause the IHS to: receive, from a workspace orchestration service, orchestration code; use a trusted controller coupled to the processor to record a log, the log including: the orchestration code, and an indication of a sequence of operations performed by a local management agent during instantiation of a workspace; provide, to the workspace orchestration service, a copy of the log; and in response to the workspace orchestration service successfully: (i) authenticating the orchestration code, and (ii) verifying the sequence of operations, establish a connection between the workspace and the workspace orchestration service.

The present disclosure provides a multi-agent simulation system and a multi-agent simulation method. A multi-agent simulation (MAS) system capable of maintaining the precision of simulation and performing simulation at high speed is provided. The MAS system (100) is provided with an agent simulator (200A to 200C) set for each agent and a central controller (300). The agent simulator (200A to 200C) simulates the state of each agent while causing the agents to interact with each other through the exchange of messages. The central controller (300) manages the participation and disengagement of the agent simulator (200A to 200C) from the simulation of the target world, and causes the agent simulator (200B) that is unable to keep up with the flow of time of the target world to disengage from the simulation of the target world.

The embodiment of the invention provides a data processing method and a multi-agent processing system, and the method comprises the steps: determining a target analysis task, and generating a target user file corresponding to the target analysis task through a user agent; according to the target analysis task, the task data and the target user file, utilizing the management agent to generate a task execution plan corresponding to the target analysis task; generating a verification analysis plan and a verification subtask corresponding to the verification analysis plan according to the task execution plan by using the management agent and the analysis model; and executing the verification subtask corresponding to the verification analysis plan by using the management agent to obtain a task execution result, and generating a task analysis report corresponding to the target analysis task according to the task execution result. Through interaction cooperation among multiple agents, the analysis task can be autonomously planned when the target analysis task is processed, the attribution hypothesis is dynamically generated and verified, and a high-value task analysis report is generated.

This invention relates to a multi-agent collaborative management method and system for water data governance, belonging to the field of water data governance technology. It includes: constructing a data asset management agent, a quality diagnosis agent, and a conversational analysis agent, and setting a standardized decision-making process; building a structured message queue; when water demand processing data is received, the data asset management agent analyzes the water demand processing data to generate water management data and sends it to the message queue; when the quality diagnosis agent listens for water management data in the message queue, it retrieves the water management data, analyzes it, generates governance strategy instructions, and sends them to the message queue; combining the conversational analysis agent and the governance strategy instructions, it determines governance adjustment instructions and executes the governance adjustment instructions. This invention improves the overall efficiency of water data governance and shortens the average problem resolution cycle.

The invention discloses a recommendation method, device and system, and belongs to the technical field of artificial intelligence. The method is applied to a recommendation system comprising a management agent and at least one execution type agent, the method is executed by the management agent, and the method comprises the steps of obtaining a recommendation task; determining at least one subtask of the recommended task; calling an execution type intelligent agent used for executing the at least one subtask in at least one execution type intelligent agent included in the recommendation system to execute the at least one subtask so as to obtain a task result of the at least one subtask; and determining a recommendation result for the recommendation task according to the task demand of the recommendation task and the task result of the at least one subtask. According to the method, the recommendation task is cooperatively processed by using multiple intelligent agents with different capabilities, so that the recommendation accuracy and efficiency are improved.

Implementations for management resilience with route persistence and state reconciliation in an accelerator interconnect network. In some implementations, a system comprises accelerators coupled to switches via protocol-level interface links, the switches storing forwarding entries in hardware, each forwarding entry mapping a destination accelerator identifier to an egress port. A centralized controller generates and distributes the forwarding entries via management agents. The forwarding entries persist in the switches independently of the centralized controller, and when the controller fails, the switches continue routing transactions and executing workloads continue. In some implementations, upon recovery, the centralized controller re-establishes connections to management agents, re-gathers identity, configuration, and operational state, identifies discrepancies between intended and actual configuration, pushes corrections, and recalculates forwarding entries. The reconciliation may use differential comparison with cached pre-failure state and may prioritize data plane correctness.

Provided is an information processing system for an autonomous system, the information processing system obtaining observation sensor data in the autonomous system and a target of the autonomous system, and including: an agent management unit that stores and manages an agent group; an attention unit configured to select an agent from the agent group based on the target; and an agent cooperation unit configured to generate a new agent in which inputs and outputs of selected agents are integrated and to train a generative model of the new agent based on the observation sensor data.

Embodiments of the present application provide a power system network security protection method and device based on multiple agents. The method comprises: when a detection agent detects a real-time network attack, obtaining attack event information reported by the detection agent; issuing the attack event information to a tracing agent to trigger the tracing agent to obtain a target tracing investigation model from a beacon management agent; controlling the tracing agent to use the target tracing investigation model to conduct a tracing investigation on the real-time network attack to determine a target portrait corresponding to the real-time network attack; based on the target portrait, using a red teaming agent to generate at least one coping strategy; determining a target strategy in the at least one coping strategy, and using the red teaming agent to execute the target strategy. The method is used to improve the response efficiency to network attacks and ensure the safe and stable operation of the power system.

A large language model (LLM) twin system can include a user interface to receive a query from a user, a Digital Twin to simulate and verify parameters of each of a plurality of devices of a factory, and a plurality of LLM Agents including a Manager Agent for controlling and communicating with the other LLM Agents and for receiving the query, interpreting the query using one or more of the LLM Agents to update the query, sending the update query to one or more of the other LLM Agents to receive an answer to the query, receiving the answer from the one or more other LLM Agents, and sending the answer to the user through the user interface.

The application relates to a multi-agent collaborative clinical scientific research implementation method and device and computer equipment. The method comprises the following steps: in response to a user's clinical scientific research text input operation, generating demand information, obtaining a search instruction through dialogue interaction hub analysis, and obtaining a first preset number of candidate data sets through a medical data set management agent; after the user selects, a training data set is determined, a data set instruction is obtained by combining the search instruction and the training data set, a second preset number of candidate algorithms are generated by an algorithm management agent, the user selects and determines a training algorithm and generates a configuration file; based on the training algorithm and the configuration file, a third preset number of candidate algorithm power schemes are generated through an algorithm power scheduling module, the user selects and determines a training scheme and generates a training configuration file; a clinical scientific research algorithm is obtained through iterative training based on the file, an inference configuration file is generated in response to an inference text input, and finally an inference report is generated. The method can automatically implement clinical scientific research.