DCG-based agent orchestration and task scheduling method and device

Abstract

The present application provides a method and device for intelligent agent orchestration and task scheduling based on DCG, which relates to the field of artificial intelligence technology and solves the technical problem that the existing task scheduling method makes task management less flexible. The method includes: obtaining multiple task nodes; determining the dependency relationship between the multiple task nodes based on the directed edges in the DCG, and controlling the execution order of the multiple task nodes based on the dependency relationship so that the tasks corresponding to the task nodes are executed in the order corresponding to the dependency relationship; dynamically allocating multiple tasks to corresponding suitable intelligent agents for execution based on the dependency relationship and the current state of the intelligent agent; controlling the parallel execution of multiple tasks based on the dependency relationship and the load of the intelligent agent to realize the execution of the workflow; and dynamically adjusting the execution path of the task based on the real-time monitoring data of the task during the execution of the workflow to realize the real-time allocation of the task.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a DCG-based intelligent agent orchestration and task scheduling method and device. Background Art

[0002] Task scheduling is a crucial component of operating systems, and for real-time operating systems, task scheduling directly impacts their real-time performance. Task scheduling algorithms can be categorized as event-driven or clock-driven. Event-driven scheduling schedules tasks based on the sequence of events and their priority; clock-driven scheduling is generally used for periodic tasks.

[0003] Currently, existing task scheduling and management methods usually rely on linear or static methods to handle task dependencies. As a result, the static scheduling method of the task scheduling system leads to low task management flexibility and difficulty in coping with dynamic adjustment requirements in complex scenarios. Summary of the Invention

[0004] The purpose of the present invention is to provide a method and device for agent orchestration and task scheduling based on a directed cyclic graph (DCG) to solve the technical problem that the existing task scheduling method has low task management flexibility.

[0005] In a first aspect, the present application provides a DCG-based agent orchestration and task scheduling method, the method comprising:

[0006] Acquire a plurality of task nodes, wherein each task node represents a task performed by an agent, the plurality of task nodes and the plurality of agents are connected by directed edges, and a directed cyclic graph (DCG) is formed by the plurality of task nodes, the plurality of agents, and the directed edges;

[0007] Determining dependency relationships between the plurality of task nodes according to the directed edges in the DCG, and controlling the execution order of the plurality of task nodes according to the dependency relationships so that the tasks corresponding to the task nodes are executed in an order corresponding to the dependency relationships;

[0008] Dynamically assigning multiple tasks to corresponding suitable agents for execution based on the dependency relationships and the current states of the agents; wherein the multiple agents collaborate to complete the tasks through a feedback mechanism;

[0009] According to the dependency relationship and the load of the agent, controlling the parallel execution of the plurality of tasks to realize the execution of the workflow;

[0010] During the execution of the workflow, the execution path of the task is dynamically adjusted according to the data of real-time monitoring of the task, so as to achieve real-time allocation of the task.

[0011] In one possible implementation, the DCG formed by the plurality of task nodes and the directed edges represents a dependency relationship between the plurality of task nodes; and controlling the execution order of the plurality of task nodes according to the dependency relationship so that the tasks corresponding to the task nodes are executed in an order corresponding to the dependency relationship includes:

[0012] According to the dependency relationship, the execution order of the plurality of task nodes is controlled by a topological sorting algorithm in a graph structure, so that the tasks corresponding to the task nodes are executed in the order corresponding to the dependency relationship.

[0013] In one possible implementation, dynamically allocating the plurality of tasks to corresponding suitable agents for execution based on the dependency relationship and the current state of the agent includes:

[0014] According to the task priority corresponding to the dependency relationship, the execution capability of the agent and the load of the agent, a suitable agent corresponding to the task is dynamically selected from multiple agents, and the task is assigned to the suitable agent for execution.

[0015] In one possible implementation, controlling the parallel execution of the plurality of tasks according to the dependency relationship and the load of the agent includes:

[0016] Automatically determine a parallel target task node among the plurality of task nodes through a graph structure.

[0017] According to the dependency relationship and the load condition of the agent, the target tasks corresponding to the plurality of target task nodes are controlled to be executed in parallel, and the agent is assigned to execute the plurality of target tasks simultaneously.

[0018] In one possible implementation, dynamically adjusting the execution path of the task based on data from real-time monitoring of the task includes:

[0019] The execution progress of the task is monitored in real time, and the execution path of the task is dynamically adjusted according to the execution progress fed back by the task node to handle delays or blockages that occur during the execution of the task.

[0020] In one possible implementation, the agent's orchestration framework performs dependency management and dynamic scheduling between the tasks through a graph structure, and optimizes the execution path of the tasks based on real-time data of the execution progress to achieve system responsiveness.

[0021] In one possible implementation, the agent orchestration framework is used to manage the dependencies between the multiple tasks, and to dynamically schedule and execute the tasks in parallel under the collaborative environment of the multiple agents.

[0022] In a possible implementation, the plurality of task nodes include any one or more of the following: a summary task node, a search task node, and an extraction task node.

[0023] In a second aspect, the present application provides a DCG-based agent orchestration and task scheduling device, the device comprising:

[0024] An acquisition module is configured to acquire a plurality of task nodes, wherein each task node represents a task performed by an agent, the plurality of task nodes and the plurality of agents are connected by directed edges, and a directed cyclic graph (DCG) is formed by the plurality of task nodes, the plurality of agents, and the directed edges;

[0025] A dependency control module is configured to determine the dependency relationships between the plurality of task nodes according to the directed edges in the DCG, and control the execution order of the plurality of task nodes according to the dependency relationships, so that the tasks corresponding to the task nodes are executed in the order corresponding to the dependency relationships;

[0026] A dynamic task allocation module is used to dynamically allocate multiple tasks to corresponding suitable agents for execution based on the dependency relationship and the current state of the agent; wherein the multiple agents collaborate to complete the tasks through a feedback mechanism;

[0027] A parallel execution module, configured to control the parallel execution of the plurality of tasks according to the dependency relationship and the load of the agent, so as to realize the execution of the workflow;

[0028] The real-time adjustment module is used to dynamically adjust the execution path of the task according to the data of real-time monitoring of the task during the execution of the workflow, so as to achieve real-time allocation of the task.

[0029] In a third aspect, the present application provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, the method described in the first aspect is implemented.

[0030] In a fourth aspect, the present application further provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are called and executed by a processor, the computer-executable instructions prompt the processor to execute the method described in the first aspect above.

[0031] This application brings the following beneficial effects:

[0032] The present application provides a DCG-based agent orchestration and task scheduling method and device, which can obtain multiple task nodes; wherein each task node represents a task performed by an agent, and multiple task nodes and multiple agents are connected by directed edges, and a directed cyclic graph DCG is formed by multiple task nodes, multiple agents and the directed edges; the dependency relationship between the multiple task nodes is determined according to the directed edges in the DCG, and the execution order of the multiple task nodes is controlled according to the dependency relationship, so that the tasks corresponding to the task nodes are executed in the order corresponding to the dependency relationship; according to the dependency relationship and the current state of the agent, the multiple tasks are dynamically allocated to the corresponding suitable agents for execution; wherein the multiple agents collaborate to complete the tasks through a feedback mechanism; according to the dependency relationship and the load condition of the agent, the multiple tasks are controlled to be executed in parallel to realize the execution of the workflow; during the execution of the workflow, the execution path of the task is dynamically adjusted according to the real-time monitoring data of the task to realize the real-time allocation of the task. In this scheme, the dependency relationship between tasks and agents is represented by the graph structure of directed cyclic graph (DCG), which supports dynamic scheduling and path selection of tasks. Moreover, the system dynamically adjusts the task order and selects the optimal execution path according to the task progress, agent status and real-time data, thereby avoiding fixed-order execution and improving execution efficiency and resource utilization. At the same time, the system has a real-time feedback mechanism that can optimize task paths and allocations according to task execution status. Therefore, this method can flexibly handle complex dependencies between tasks, efficiently manage task dependencies, support parallel execution and collaboration between agents, and realize dynamic scheduling of the task scheduling system through collaboration and dynamic adjustment between agents, thereby improving the flexibility of task management and the execution efficiency and accuracy of the workflow to meet the dynamic adjustment needs in complex scenarios, solving the technical problem that the existing task scheduling method makes task management less flexible.

[0033] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the specific implementation methods of the present application or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the specific implementation methods or the description of the prior art. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0035] Figure 1 A schematic diagram of the process of the DCG-based agent orchestration and task scheduling method provided in an embodiment of the present application;

[0036] Figure 2 An example of a workflow structure diagram in the DCG-based agent orchestration and task scheduling method provided in an embodiment of the present application;

[0037] Figure 3 An example of a flowchart for task execution path selection in the DCG-based agent orchestration and task scheduling method provided in an embodiment of the present application;

[0038] Figure 4 A schematic diagram of the structure of a DCG-based agent orchestration and task scheduling device provided in an embodiment of the present application;

[0039] Figure 5 A schematic structural diagram of an electronic device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0040] To make the purpose, technical solutions, and advantages of the embodiments of this application more clear, the technical solutions of this application will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0041] The terms "including," "having," and any variations thereof, as used in the embodiments of this application, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or units is not limited to the listed steps or units but may optionally include other steps or units not listed, or may optionally include other steps or units inherent to the process, method, product, or apparatus.

[0042] Currently, traditional task scheduling and management systems lack flexibility and real-time adjustment capabilities, especially in multi-agent collaborative environments. They are unable to efficiently allocate tasks and manage inter-task dependencies. Furthermore, existing workflow management systems often suffer from low task execution efficiency, slow response times, and an inability to dynamically adapt to task changes. Furthermore, existing task management systems lack the ability to handle complex task dependencies, dynamic scheduling, and parallel execution.

[0043] Based on this, an embodiment of the present application provides a DCG-based intelligent agent orchestration and task scheduling method and device, which can solve technical problems such as the low flexibility of task management in existing task scheduling methods.

[0044] The embodiments of the present invention are further described below with reference to the accompanying drawings.

[0045] Figure 1 The following is a flow chart of a DCG-based agent orchestration and task scheduling method provided in an embodiment of the present application. Figure 1 As shown, the method includes:

[0046] Step S110: Acquire multiple task nodes.

[0047] Each task node represents a task performed by an agent. Multiple task nodes and multiple agents are connected by directed edges. A directed cyclic graph (DCG) is formed by multiple task nodes, multiple agents, and directed edges.

[0048] As a possible implementation, Figure 2 As shown, the node searchnode is searched in the task node steps node, then the node extract node is extracted from the searched node, and finally the summary node summarize node is executed based on the extracted node; of course, the summary node summarize node can also be directly summarized after the task node steps node.

[0049] In some embodiments, the plurality of task nodes include any one or more of the following: a summary task node, a search task node, and an extraction task node. In the embodiment of the present application, the task nodes include a summary task node, a search task node, and an extraction task node, and the task nodes are connected by directed edges to form a directed cyclic graph, which represents the dependency relationship between tasks.

[0050] As an optional implementation, taking the intelligent tour guide system for scenic spots as an example, tour tasks are modeled based on a DCG graph structure, and combined with an agent collaboration mechanism to achieve dynamic scheduling and parallel task execution. For example, during the task modeling process, each scenic spot, activity, and explanation content is modeled as a task node, and the dependencies between tasks are represented by a directed cyclic graph. For example, at a certain scenic spot, tourists may visit, take photos, and listen to the tour in parallel, but there may also be sequential dependencies, such as when visiting a historical site, they need to listen to the explanation first and then engage in interactive activities.

[0051] Step S120 , determining the dependency relationships among multiple task nodes based on the directed edges in the DCG, and controlling the execution order of the multiple task nodes based on the dependency relationships, so that the tasks corresponding to the task nodes are executed in the order corresponding to the dependency relationships.

[0052] As a possible implementation, the dependency control module is used to control the execution order of tasks based on the defined dependencies between task nodes, ensuring that tasks are executed in the order of the dependencies. It should be noted that the directed edges between task nodes and agent nodes represent the dependencies between tasks.

[0053] In some embodiments, a DCG formed by multiple task nodes and directed edges represents the dependency relationship between multiple task nodes; the above-mentioned control of the execution order of multiple task nodes based on the dependency relationship so that the tasks corresponding to the task nodes are executed in the order corresponding to the dependency relationship can specifically include the following steps: controlling the execution order of multiple task nodes through a topological sorting algorithm in the graph structure based on the dependency relationship so that the tasks corresponding to the task nodes are executed in the order corresponding to the dependency relationship. For example, the dependency control module ensures that the task nodes are executed in the order of the task dependencies through the topological sorting algorithm in the graph structure, thereby avoiding dependency conflicts caused by errors in the task execution order.

[0054] In step S130 , multiple tasks are dynamically assigned to corresponding suitable agents for execution based on the dependency relationship and the current state of the agent.

[0055] Multiple agents collaborate to complete tasks through a feedback mechanism. As an optional implementation, a dynamic task allocation module dynamically assigns tasks to appropriate agents based on the dependencies between task nodes and the agent status. Agents collaborate to complete tasks through a feedback mechanism. This multi-agent collaboration allows for flexible adjustment of task load distribution and execution paths.

[0056] For example, in multi-agent collaboration and task allocation, multiple agents (such as virtual tour guides, smart devices, and human guides) work together to dynamically allocate tasks based on real-time feedback and visitor needs. This inter-agent collaboration ensures efficient task completion and adjusts task execution strategies based on visitor needs (such as interests, language, and group size).

[0057] In some embodiments, this step may specifically include the following steps: dynamically selecting an appropriate agent corresponding to the task from multiple agents based on the task priority corresponding to the dependency, the agent's execution capability, and the agent's load, and assigning the task to the appropriate agent for execution. Exemplarily, the dynamic task assignment module dynamically selects an appropriate agent to perform the task based on each agent's execution capability, load, and task priority, thereby improving the efficiency and accuracy of task execution.

[0058] Step S140 , controlling the parallel execution of multiple tasks based on the dependency relationship and the load of the agent to implement the execution of the workflow.

[0059] In the embodiment of the present application, the parallel execution module supports the parallel execution of multiple tasks based on the dependencies between task nodes and the load of the agent, thereby improving the efficiency of workflow execution. For example, in terms of parallel execution of tasks and resource optimization, the system can execute multiple tasks simultaneously, determine which tasks can be executed in parallel through the topological structure of the DCG graph, and ensure the efficient execution of parallel tasks through agent collaboration. At the same time, resources (such as equipment, tour guides, etc.) are dynamically scheduled according to the real-time load to avoid conflicts and delays.

[0060] In some embodiments, this step may specifically include the following steps: automatically determining parallel target task nodes among multiple task nodes through a graph structure; controlling the parallel execution of target tasks corresponding to multiple target task nodes based on dependency relationships and the load of the intelligent agent, and allocating the intelligent agent to execute multiple target tasks simultaneously.

[0061] For example, the parallel execution module supports the parallel execution of multiple task nodes. Agents can perform multiple tasks simultaneously, automatically identifying and assigning parallelizable task nodes through the graph structure. To meet the needs of complex scenarios, the flexible modeling of the graph structure enables the system to handle complex task dependencies and support multi-agent collaboration, thus adapting to different task types and complex workflow scenarios.

[0062] In some embodiments, the agent orchestration framework is used to manage the dependencies between multiple tasks and dynamically schedule and execute tasks in parallel in a collaborative environment. The agent orchestration framework can efficiently manage the complex dependencies between multiple tasks, dynamically schedule and execute tasks in parallel in a collaborative environment, and optimize the execution path based on real-time feedback to ensure efficient and accurate task execution. Furthermore, through dynamic scheduling and parallel execution, the system can optimize task allocation in real time, reduce task execution time, and thus improve the overall efficiency of the workflow.

[0063] Step S150 , during the execution of the workflow, dynamically adjust the execution path of the task according to the data of real-time monitoring of the task to achieve real-time distribution of the task.

[0064] As one possible implementation, a real-time adjustment module dynamically adjusts task execution paths during workflow execution based on real-time monitoring data (i.e., the monitored task execution progress) to optimize task allocation. For example, in terms of real-time dynamic task scheduling, the system can adjust task execution paths in real time based on visitor behavior, interests, and scenic area environment (such as weather and visitor flow). For example, if a particular attraction is overcrowded, the system can automatically adjust the visitor's route to avoid congestion.

[0065] In some embodiments, this step may specifically include the following steps: real-time monitoring of the execution progress of the task, and dynamically adjusting the execution path of the task according to the execution progress feedback from the task node to handle delays or blockages that occur during the execution of the task. Figure 3 As shown, after checking task progress, the system determines whether the task is complete. If yes, it proceeds to the next task (Move to next task); if no, it waits for the task to complete (wait for task to complete). As an optional implementation, a real-time adjustment module monitors task execution progress in real time and dynamically adjusts the task execution path based on feedback from task nodes to address potential delays or blockages during task execution.

[0066] In terms of real-time feedback and route selection optimization, the system can dynamically adjust task paths based on real-time feedback data to avoid task delays, optimize execution order, and ensure timely task completion. In particular, when faced with task delays or agent unavailability, the system can immediately react and adjust the path to avoid wasted resources.

[0067] The following is an example of the implementation of the intelligent guide system in the Zigong Dinosaur Museum. The Zigong Dinosaur Museum's intelligent guide system dynamically schedules according to the real-time needs of tourists, the priority of tasks, and the status of intelligent agents, and selects the optimal path through the topological structure of the graph to achieve parallel execution of tasks and flexible allocation of intelligent agents. Each exhibit and display area corresponds to an intelligent agent node in the system, and the intelligent agent is responsible for providing detailed explanations or interactive experiences to tourists. When tourists enter the museum, their visit progress and points of interest will be recorded in real time, and the intelligent agent will adjust the explanation content or recommend other relevant exhibits based on the tourists' feedback and needs. By tracking the location and progress of tourists in real time, the system can obtain data such as the flow of tourists in the museum and the visiting conditions of exhibits in real time, dynamically adjust the tourists' visiting routes, and intelligently select the optimal path, thereby improving the efficiency and comfort of tourists' visits. The method provided in the embodiment of the present application can be widely applied to the fields of intelligent guide in scenic spots, automated task management, smart tourism services, etc., and can significantly improve the efficiency, flexibility and system responsiveness of task execution.

[0068] In some embodiments, the orchestration framework of the intelligent agent performs dependency management and dynamic scheduling between tasks through a graph structure, and optimizes the execution path of the task according to the real-time data of the execution progress to achieve system response. Among them, the orchestration framework of the intelligent agent realizes dependency management and dynamic scheduling between tasks through a graph structure, and can optimize the task execution path according to real-time data (the execution progress of the above tasks), improve the system response speed, and solve the problem that the task execution order in the traditional task management system is fixed and cannot be flexibly adjusted. Moreover, unlike traditional linear or static task scheduling, the dynamic adjustment capability of this solution enables the system to respond to sudden changes and optimize the execution path in a timely manner, thereby ensuring the efficient operation of the system in a complex environment and improving the flexibility of task management.

[0069] The dependency relationship between tasks and agents is represented by the graph structure of a directed cyclic graph (DCG), which supports dynamic scheduling and path selection of tasks. Moreover, the system dynamically adjusts the task sequence and selects the optimal execution path based on task progress, agent status and real-time data, thereby avoiding fixed-order execution and improving execution efficiency and resource utilization. At the same time, the system has a real-time feedback mechanism that can optimize task paths and allocations based on task execution status. Therefore, this method can flexibly handle complex dependencies between tasks, efficiently manage task dependencies, support parallel execution and collaboration between agents, and realize dynamic scheduling of the task scheduling system through collaboration and dynamic adjustment between agents, thereby improving the flexibility of task management to meet the dynamic adjustment needs in complex scenarios.

[0070] Moreover, the orchestration framework of the intelligent agent improves the execution efficiency of the workflow through the DCG graph structure and the collaboration of intelligent agents. It is particularly suitable for multi-task management and dynamic scheduling requirements in complex environments, solves the defects in the existing technology, and improves the execution efficiency and accuracy of the workflow. Specifically, the orchestration framework of the intelligent agent constructs a DCG-based task scheduling structure through a graph structure, and represents the dependency relationship of the task through the directed edges between the task nodes and the intelligent agent nodes; secondly, the execution order of the tasks is calculated using topological sorting to ensure that the task dependencies are met; then, tasks are dynamically allocated and executed in parallel according to the status, resource conditions and priority of the intelligent agent; then, a real-time feedback mechanism is used to dynamically select the task execution path to avoid the limitations of fixed-order execution; finally, the intelligent agent completes the task through collaboration and adjusts the task path in time to ensure the smooth completion of the task. Moreover, the task management framework in the embodiment of the present application can be dynamically scheduled according to the task dependency relationship, and tasks can be efficiently completed through collaboration between intelligent agents.

[0071] In this application, a multi-agent orchestration framework based on a directed cyclic graph (DCG) is provided, leveraging task scheduling and execution path selection techniques within a multi-agent system. This framework is used to flexibly and efficiently manage task scheduling and agent collaboration within intelligent scenic area navigation. This approach dynamically adjusts task scheduling and execution paths, and is widely applicable to intelligent scenic area navigation, automated task management, smart tourism services, and other fields.

[0072] Figure 4 A schematic diagram of the structure of a DCG-based agent orchestration and task scheduling device is provided. Figure 4 As shown, the DCG-based agent orchestration and task scheduling device 400 includes:

[0073] An acquisition module 401 is configured to acquire a plurality of task nodes, wherein each task node represents a task performed by an agent, the plurality of task nodes and the plurality of agents are connected by directed edges, and a directed cyclic graph (DCG) is formed by the plurality of task nodes, the plurality of agents, and the directed edges;

[0074] A dependency control module 402 is configured to determine the dependency relationships between the plurality of task nodes based on the directed edges in the DCG, and control the execution order of the plurality of task nodes based on the dependency relationships so that the tasks corresponding to the task nodes are executed in the order corresponding to the dependency relationships;

[0075] A dynamic task allocation module 403 is configured to dynamically allocate the plurality of tasks to the corresponding suitable agents for execution based on the dependency relationships and the current states of the agents; wherein the plurality of agents collaborate to complete the tasks through a feedback mechanism;

[0076] A parallel execution module 404 is used to control the parallel execution of multiple tasks according to the dependency relationship and the load of the agent to realize the execution of the workflow;

[0077] The real-time adjustment module 405 is used to dynamically adjust the execution path of the task according to the data of real-time monitoring of the task during the execution of the workflow, so as to achieve real-time allocation of the task.

[0078] The DCG-based intelligent agent orchestration and task scheduling device provided in the embodiment of the present application has the same technical features as the DCG-based intelligent agent orchestration and task scheduling method provided in the above embodiment, so it can also solve the same technical problems and achieve the same technical effects.

[0079] An electronic device provided in an embodiment of the present application is Figure 5 As shown, the electronic device 500 includes a processor 502 and a memory 501 , wherein the memory stores a computer program that can be run on the processor, and the processor implements the steps of the method provided in the above embodiment when executing the computer program.

[0080] See also Figure 5 The electronic device further includes: a bus 503 and a communication interface 504, a processor 502, a communication interface 504 and a memory 501 connected via the bus 503; the processor 502 is used to execute executable modules stored in the memory 501, such as computer programs.

[0081] The memory 501 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage. The system network element communicates with at least one other network element via at least one communication interface 504 (which may be wired or wireless), and may utilize the Internet, a wide area network, a local area network, a metropolitan area network, or the like.

[0082] The bus 503 may be an ISA bus, a PCI bus, or an EISA bus. The bus may be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, Figure 5 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0083] Among them, the memory 501 is used to store programs, and the processor 502 executes the program after receiving the execution instruction. The method executed by the process definition device disclosed in any embodiment of the present application can be applied to the processor 502 or implemented by the processor 502.

[0084] The processor 502 may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by an integrated logic circuit of hardware in the processor 502 or by instructions in the form of software. The above-mentioned processor 502 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The various methods, steps, and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in conjunction with the embodiments of the present application can be directly embodied as being executed by a hardware decoding processor, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium well-known in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in memory 501, and processor 502 reads the information in memory 501 and, in conjunction with its hardware, completes the steps of the above method.

[0085] Corresponding to the above-mentioned DCG-based intelligent agent orchestration and task scheduling method, an embodiment of the present application also provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are called and executed by the processor, the computer-executable instructions prompt the processor to execute the steps of the above-mentioned DCG-based intelligent agent orchestration and task scheduling method.

[0086] The DCG-based intelligent agent orchestration and task scheduling device provided in the embodiment of the present application can be specific hardware on the device or software or firmware installed on the device. The implementation principle and technical effects of the device provided in the embodiment of the present application are the same as those of the aforementioned method embodiment. For the sake of brief description, for matters not mentioned in the device embodiment, reference can be made to the corresponding content in the aforementioned method embodiment. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can all refer to the corresponding processes in the aforementioned method embodiment, and will not be repeated here.

[0087] In the embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interface, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0088] For another example, the flowcharts and block diagrams in the accompanying drawings show the possible architectures, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of code, and the module, program segment or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of the boxes in the block diagram and / or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0089] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0090] In addition, each functional unit in the embodiments provided in the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0091] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the DCG-based intelligent agent orchestration and task scheduling method described in each embodiment of the present application. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, and other media that can store program codes.

[0092] It should be noted that similar numbers and letters represent similar items in the following figures. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. In addition, the terms "first", "second", "third", etc. are only used to distinguish the description and are not to be understood as indicating or implying relative importance.

[0093] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The scope of protection of the present application is not limited thereto. Although the present application has been described in detail with reference to the above-mentioned embodiments, those skilled in the art should understand that any person skilled in the art can modify or easily conceive of changes to the technical solutions described in the above-mentioned embodiments within the technical scope disclosed in the present application, or perform equivalent replacements for some of the technical features thereof. However, these modifications, changes, or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application. They should all be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A DCG-based agent orchestration and task scheduling method, characterized in that: The method comprises: Acquire multiple task nodes; wherein each task node represents a task performed by an agent, the multiple task nodes and the multiple agents are connected by directed edges, and a directed cyclic graph (DCG) is formed by the multiple task nodes, the multiple agents, and the directed edges; Determining dependency relationships between the plurality of task nodes according to the directed edges in the DCG, and controlling the execution order of the plurality of task nodes according to the dependency relationships so that the tasks corresponding to the task nodes are executed in an order corresponding to the dependency relationships; Dynamically assigning multiple tasks to corresponding suitable agents for execution based on the dependency relationships and the current states of the agents; wherein the multiple agents collaborate to complete the tasks through a feedback mechanism; According to the dependency relationship and the load of the agent, controlling the parallel execution of the plurality of tasks to realize the execution of the workflow; During the execution of the workflow, the execution path of the task is dynamically adjusted according to the data of real-time monitoring of the task, so as to achieve real-time allocation of the task.

2. The method according to claim 1, characterized in that The DCG formed by the plurality of task nodes and the directed edges represents the dependency relationship between the plurality of task nodes; and controlling the execution order of the plurality of task nodes according to the dependency relationship so that the tasks corresponding to the task nodes are executed in the order corresponding to the dependency relationship, including: According to the dependency relationship, the execution order of the plurality of task nodes is controlled by a topological sorting algorithm in a graph structure, so that the tasks corresponding to the task nodes are executed in the order corresponding to the dependency relationship.

3. The method according to claim 1, characterized in that The dynamically allocating the plurality of tasks to corresponding suitable agents for execution based on the dependency relationship and the current state of the agent includes: According to the task priority corresponding to the dependency relationship, the execution capability of the agent and the load of the agent, a suitable agent corresponding to the task is dynamically selected from multiple agents, and the task is assigned to the suitable agent for execution.

4. The method according to claim 1, wherein The controlling the parallel execution of the plurality of tasks according to the dependency relationship and the load of the agent includes: Automatically determine a parallelizable target task node among the plurality of task nodes through a graph structure; According to the dependency relationship and the load condition of the agent, the target tasks corresponding to the plurality of target task nodes are controlled to be executed in parallel, and the agent is assigned to execute the plurality of target tasks simultaneously.

5. The method according to claim 2, characterized in that The dynamically adjusting the execution path of the task according to the data of real-time monitoring of the task includes: The execution progress of the task is monitored in real time, and the execution path of the task is dynamically adjusted according to the execution progress fed back by the task node to handle delays or blockages that occur during the execution of the task.

6. The method according to claim 5, characterized in that The orchestration framework of the intelligent agent performs dependency management and dynamic scheduling between the tasks through a graph structure, and optimizes the execution path of the tasks according to the real-time data of the execution progress to achieve system response.

7. The method according to claim 6, characterized in that The agent orchestration framework is used to manage the dependencies between the multiple tasks and to dynamically schedule and execute the tasks in parallel under the collaborative environment of the multiple agents.

8. The method according to claim 1, characterized in that The multiple task nodes include any one or more of the following: a summary task node, a search task node, and an extraction task node.

9. A DCG-based agent orchestration and task scheduling device, characterized in that: The device comprises: An acquisition module is configured to acquire a plurality of task nodes, wherein each task node represents a task performed by an agent, the plurality of task nodes and the plurality of agents are connected by directed edges, and a directed cyclic graph (DCG) is formed by the plurality of task nodes, the plurality of agents, and the directed edges; A dependency control module is configured to determine the dependency relationships between the plurality of task nodes according to the directed edges in the DCG, and control the execution order of the plurality of task nodes according to the dependency relationships, so that the tasks corresponding to the task nodes are executed in the order corresponding to the dependency relationships; A dynamic task allocation module is used to dynamically allocate multiple tasks to corresponding suitable agents for execution based on the dependency relationship and the current state of the agent; wherein the multiple agents collaborate to complete the tasks through a feedback mechanism; A parallel execution module, configured to control the parallel execution of the plurality of tasks according to the dependency relationship and the load of the agent, so as to realize the execution of the workflow; The real-time adjustment module is used to dynamically adjust the execution path of the task according to the data of real-time monitoring of the task during the execution of the workflow, so as to achieve real-time allocation of the task.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are called and executed by a processor, the computer-executable instructions prompt the processor to execute the method according to any one of claims 1 to 8.

Images