Multi-task reasoning scheduling method and device

Abstract

The invention relates to the technical field of task scheduling, and discloses a multi-task reasoning scheduling method and device.The multi-task reasoning scheduling method comprises the following steps that S1, initialization is conducted, specifically, parameters of a system are subjected to initialization setting; s2, task creation: creating tasks and setting priority parameters of the tasks; s3, performing resource evaluation; s4, sorting the tasks; s5, resource allocation; s6, resource scheduling; s7, task execution: executing a task command by utilizing the allocated resources; s8, tracking and monitoring: monitoring task execution and resource utilization conditions; s9, executing feedback; s10, performing information synchronization; s11, performing exception handling; s12, performing performance evaluation; and S13, mechanism adjustment. The method is reasonable in design, the utilization efficiency of computing resources can be remarkably improved, task execution delay is reduced, and the throughput of the system is improved. Meanwhile, through continuous learning and optimization, the system can adapt to constantly changing task types and resource conditions, and long-term stable and efficient operation is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of task scheduling, and in particular to a multi-task reasoning scheduling method and device. Background Art

[0002] A thread, also known as a lightweight process, is an entity in a process. A thread itself does not own system resources, but it can share all resources owned by the process with other threads belonging to the same process. A thread can create and revoke another thread, and multiple threads in the same process can execute concurrently. Threads can switch between each other, but the time overhead of switching between threads is very large. Therefore, when the number of threads is large, the time overhead caused by thread switching will be very large, which will affect the performance of the system. Developers cannot fully anticipate the demand for products in the early stages of development. Generally, the PC side does not need to consider energy consumption and power supply issues, so it is generally operated in maximum performance mode. Therefore, although the thread overhead is large, it basically does not affect the performance of the PC system.

[0003] With the rapid development of artificial intelligence and computing power, multitasking has become an indispensable part of modern computing systems. When multiple tasks need to share the same computing resources, how to effectively schedule these tasks to optimize resource utilization, improve task execution efficiency, reduce energy consumption, and ensure task fairness and real-time performance becomes a key issue; therefore, we propose a multitasking reasoning scheduling method and device to solve this problem. Summary of the invention

[0004] The purpose of the present invention is to solve the shortcomings mentioned in the background technology and to propose a multi-task reasoning scheduling method and device.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A multi-task reasoning scheduling method comprises the following steps:

[0007] S1. Initialization: Initialize the system parameters;

[0008] S2, Task creation: Create a task and set the priority parameters of the task;

[0009] S3, Resource evaluation: Evaluate the ratio of resources occupied by tasks and system available resources;

[0010] S4, Task Sorting: Prioritize multiple tasks according to their priority parameters and resource usage ratios;

[0011] S5. Resource allocation: Allocate resources to each character according to the priority classification;

[0012] S6. Resource scheduling: Scheduling different resources according to resource allocation;

[0013] S7, Task execution: Execute the task command using the allocated resources;

[0014] S8, Tracking and monitoring: monitor task execution and resource utilization;

[0015] S9. Execution feedback: Provide timely feedback on the tracking and monitoring status;

[0016] S10, information synchronization: adaptively adjust the priority and resource allocation according to the feedback information;

[0017] S11, Exception handling: Quickly respond and adjust the execution status when an emergency interrupt command occurs;

[0018] S12. Performance evaluation: Regularly evaluate and count resource utilization, execution efficiency, and execution reliability;

[0019] S13. Mechanism adjustment: Adaptively adjust the task allocation mechanism based on the results.

[0020] Preferably, in S5, when allocating resources, attention should be paid to adopting a flexible allocation strategy to adapt to changes in demand and resource environment, and adopting an optimization algorithm to optimize resource allocation.

[0021] Preferably, in S4, when performing task priority classification, it is necessary to consider urgency, importance, dependency, task duration, resource availability, workload balance, risk-benefit and relevance requirements. It is necessary to assign higher priority to tasks with higher urgency and importance, and determine the priority ranking by considering the prerequisite relationship between tasks. Priority is assigned according to the availability of different resource types required, and higher priority is assigned to tasks with shorter duration and tasks with smaller workload balance. For high-risk tasks, it is necessary to perform a risk-benefit analysis and uniformly sort multiple related tasks according to the relevance of the tasks.

[0022] Preferably, in S6, when scheduling resources, resources of a type with a larger amount of availability are scheduled preferentially.

[0023] Preferably, the adaptive adjustment of priorities and resource allocation in S10 should ensure that key individuals can receive sufficient resource support.

[0024] The present invention also provides a multi-task reasoning scheduling device, comprising:

[0025] Distributed processor units for independent processing of multiple tasks;

[0026] Dynamic task scheduling unit, used to create, allocate and schedule resources for tasks;

[0027] Collaborative reasoning unit, used to share information and knowledge between processors and perform joint reasoning;

[0028] Resource optimization unit, used to monitor and feedback the execution of task scheduling and optimize the allocation mechanism;

[0029] The dynamic task scheduling unit is connected to the resource optimization unit, and the distributed processor unit is connected to the dynamic task scheduling unit and the collaborative reasoning unit.

[0030] Preferably, the dynamic task scheduling unit includes a task creation unit, a resource evaluation unit, a task sorting unit, a resource allocation unit and a resource scheduling module, wherein the task creation is used to create tasks and set priority parameters of tasks, the resource evaluation is used to evaluate the resource ratio occupied by tasks and system available resources, the task sorting is used to prioritize multiple tasks according to the priority parameters of the tasks and the resource ratio occupied, the resource allocation is used to allocate resources to each character according to the priority classification, and the resource scheduling is used to schedule different resources according to the resource allocation situation.

[0031] Preferably, the resource optimization unit includes a feedback module, an evaluation module and an adjustment unit. The feedback unit is used to monitor task execution and resource utilization and to feed back monitoring information. The evaluation module is used to evaluate and statistically analyze task execution and resource utilization, execution efficiency and execution reliability. The adjustment unit is used to adjust the logic of the task scheduling unit.

[0032] Preferably, the collaborative reasoning unit includes a resource collaboration module, a data collaboration module, a mechanism collaboration module and an intelligent collaboration module. The resource collaboration module is used to synchronize information and globally schedule system resources. The data collaboration module is used to collect data of distributed processor units and share them synchronously. The mechanism collaboration unit is used to synchronize the task allocation mechanism of distributed processor units. The intelligent collaboration module uses a deep learning algorithm to perform closed-loop optimization on distributed processor units.

[0033] Preferably, the distributed processor unit includes a communication module, a storage module, a processing module and a control module. The communication module is used to realize information transmission and instruction issuance between processors, the storage module is used to store, back up and restore task data and ensure the processing consistency of multiple servers, the processing module is used for task analysis, task execution and task feedback, and the control module is used to realize control of interrupt devices.

[0034] Compared with the prior art, the present invention provides a multi-task reasoning scheduling method and device, which has the following beneficial effects:

[0035] (1) Creating, allocating and scheduling tasks through a dynamic task scheduling unit. Task creation creates tasks and sets priority parameters for tasks. Resource evaluation evaluates the resource ratio occupied by tasks and the available resources of the system. Task sorting prioritizes multiple tasks according to the priority parameters and resource ratio occupied. Resource allocation allocates resources to each person according to the priority classification. Resource scheduling schedules different resources according to the resource allocation situation.

[0036] (2) monitoring and feeding back the task scheduling execution status and optimizing the allocation mechanism through the resource optimization unit, wherein the feedback unit monitors the task execution and resource utilization and feeds back the monitoring information, the evaluation module evaluates and statistics the task execution status and resource utilization, execution efficiency, and execution reliability, and the adjustment unit adjusts the logic of the task scheduling unit;

[0037] (3) Through the collaborative reasoning unit, information and knowledge are shared between processors and joint reasoning is performed. The resource collaboration module synchronizes information and globally schedules system resources. The data collaboration module collects data from distributed processor units and shares them synchronously. The mechanism collaboration unit synchronizes the task allocation mechanism of distributed processor units. The intelligent collaboration module uses deep learning algorithms to perform closed-loop optimization of distributed processor units.

[0038] The invention is reasonably designed and can significantly improve the utilization efficiency of computing resources, reduce task execution delays, and increase system throughput. At the same time, through continuous learning and optimization, the system can adapt to the ever-changing task types and resource conditions, ensuring long-term stable and efficient operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a flow chart of a multi-task reasoning scheduling method proposed by the present invention. DETAILED DESCRIPTION

[0040] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0041] In the description of the present invention, it is necessary to understand that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0042] Reference Figure 1 ,A multi-task reasoning scheduling method, ,including the following steps:

[0043] S1. Initialization: Initialize the system parameters;

[0044] S2, Task creation: Create a task and set the priority parameters of the task;

[0045] S3, Resource evaluation: Evaluate the ratio of resources occupied by tasks and system available resources;

[0046] S4, Task Sorting: Prioritize multiple tasks according to their priority parameters and resource usage ratios;

[0047] S5. Resource allocation: Allocate resources to each character according to the priority classification;

[0048] S6. Resource scheduling: Scheduling different resources according to resource allocation;

[0049] S7, Task execution: Execute the task command using the allocated resources;

[0050] S8, Tracking and monitoring: monitor task execution and resource utilization;

[0051] S9. Execution feedback: Provide timely feedback on the tracking and monitoring status;

[0052] S10, information synchronization: adaptively adjust the priority and resource allocation according to the feedback information;

[0053] S11, Exception handling: Quickly respond and adjust the execution status when an emergency interrupt command occurs;

[0054] S12. Performance evaluation: Regularly evaluate and count resource utilization, execution efficiency, and execution reliability;

[0055] S13. Mechanism adjustment: Adaptively adjust the task allocation mechanism based on the results.

[0056] In this embodiment, in S5, when allocating resources, attention should be paid to adopting a flexible allocation strategy to adapt to changes in demand and resource environment, and adopting an optimization algorithm to optimize resource allocation.

[0057] In this embodiment, in S4, when performing task priority classification, it is necessary to consider urgency, importance, dependency, task duration, resource availability, workload balance, risk-benefit and relevance requirements. It is necessary to assign higher priority to tasks with higher urgency and importance, and determine the priority ranking by considering the prerequisite relationship between tasks. Priority is assigned based on the availability of different resource types required, and higher priority is assigned to tasks with shorter duration and tasks with smaller workload balance. For high-risk tasks, a risk-benefit analysis is required, and multiple related tasks are uniformly sorted according to the relevance of the tasks.

[0058] In this embodiment, in S6, when performing resource scheduling, resources of a type with a larger amount of availability are scheduled preferentially.

[0059] In this embodiment, S10, adaptive adjustment of priorities and resource allocation should ensure that key individuals can receive sufficient resource support.

[0060] The present invention also provides a multi-task reasoning scheduling device, comprising:

[0061] Distributed processor units for independent processing of multiple tasks;

[0062] Dynamic task scheduling unit, used to create, allocate and schedule resources for tasks;

[0063] Collaborative reasoning unit, used to share information and knowledge between processors and perform joint reasoning;

[0064] Resource optimization unit, used to monitor and feedback the execution of task scheduling and optimize the allocation mechanism;

[0065] The dynamic task scheduling unit is connected to the resource optimization unit, and the distributed processor unit is connected to the dynamic task scheduling unit and the collaborative reasoning unit.

[0066] In this embodiment, the dynamic task scheduling unit includes a task creation unit, a resource evaluation unit, a task sorting unit, a resource allocation unit and a resource scheduling module. Task creation is used to create tasks and set priority parameters of tasks. Resource evaluation is used to evaluate the resource ratio occupied by tasks and system available resources. Task sorting is used to prioritize multiple tasks according to the priority parameters of the tasks and the resource occupation ratio. Resource allocation is used to allocate resources to each character according to the priority classification. Resource scheduling is used to schedule different resources according to the resource allocation situation.

[0067] In this embodiment, the resource optimization unit includes a feedback module, an evaluation module and an adjustment unit. The feedback unit is used to monitor task execution and resource utilization and to feedback monitoring information. The evaluation module is used to evaluate and compile statistics on task execution and resource utilization, execution efficiency and execution reliability. The adjustment unit is used to adjust the logic of the task scheduling unit.

[0068] In this embodiment, the collaborative reasoning unit includes a resource collaboration module, a data collaboration module, a mechanism collaboration module and an intelligent collaboration module. The resource collaboration module is used to synchronize information and globally schedule system resources. The data collaboration module is used to collect data of distributed processor units and share them synchronously. The mechanism collaboration unit is used to synchronize the task allocation mechanism of distributed processor units. The intelligent collaboration module uses deep learning algorithms to perform closed-loop optimization of distributed processor units.

[0069] In this embodiment, the distributed processor unit includes a communication module, a storage module, a processing module and a control module. The communication module is used to realize information transmission and instruction issuance between processors. The storage module is used to store, back up and restore task data and ensure the processing consistency of multiple servers. The processing module is used for task analysis, task execution and task feedback. The control module is used to realize control of interrupt devices.

[0070] In this embodiment, when in use, tasks are created, allocated and resource scheduled by a dynamic task scheduling unit. Task creation creates tasks and sets priority parameters for tasks. Resource evaluation evaluates the resource ratio occupied by tasks and system available resources. Task sorting prioritizes multiple tasks according to the priority parameters and resource ratio occupied. Resource allocation allocates resources to each person according to the priority classification. Resource scheduling schedules different resources according to the resource allocation situation.

[0071] The resource optimization unit monitors and feeds back the execution of task scheduling and optimizes the allocation mechanism, wherein the feedback unit monitors the execution of tasks and resource utilization and feeds back monitoring information, the evaluation module evaluates and counts the execution of tasks and resource utilization, execution efficiency, and execution reliability, and the adjustment unit adjusts the logic of the task scheduling unit;

[0072] Through the collaborative reasoning unit, information and knowledge are shared between processors and joint reasoning is performed. The resource collaboration module synchronizes information and globally schedules system resources. The data collaboration module collects data from distributed processor units and shares them synchronously. The mechanism collaboration unit synchronizes the task allocation mechanism of distributed processor units. The intelligent collaboration module uses deep learning algorithms to perform closed-loop optimization of distributed processor units.

[0073] Multiple tasks are processed independently through distributed processor units, the communication module realizes information transmission and instruction issuance between processors, the storage module stores, backs up and restores task data, ensures the processing consistency of multiple servers, the processing module performs task analysis, task execution and task feedback, and the control module realizes control of interrupt devices.

[0074] The standard parts used in the present invention can all be purchased from the market, and special-shaped parts can be customized according to the instructions and the drawings. The specific connection methods of each part adopt conventional means such as mature bolts, rivets, welding, etc. in the prior art. Machinery, parts and equipment all adopt conventional models in the prior art, and the circuit connection adopts the conventional connection method in the prior art, which will not be described in detail here.

Claims

1. A multi-task reasoning scheduling method, characterized in that: The steps include: S1. Initialization: Initialize the system parameters; S2, Task creation: Create a task and set the priority parameters of the task; S3, Resource evaluation: Evaluate the ratio of resources occupied by tasks and system available resources; S4, Task Sorting: Prioritize multiple tasks according to their priority parameters and resource usage ratios; S5. Resource allocation: Allocate resources to each character according to the priority classification; S6. Resource scheduling: Scheduling different resources according to resource allocation; S7, Task execution: Execute the task command using the allocated resources; S8, Tracking and monitoring: monitor task execution and resource utilization; S9. Execution feedback: Provide timely feedback on the tracking and monitoring status; S10, information synchronization: adaptively adjust the priority and resource allocation according to the feedback information; S11, Exception handling: Quickly respond and adjust the execution status when an emergency interrupt command occurs; S12. Performance evaluation: Regularly evaluate and count resource utilization, execution efficiency, and execution reliability; S13. Mechanism adjustment: Adaptively adjust the task allocation mechanism based on the results.

2. A multi-task reasoning scheduling method according to claim 1, characterized in that: In S5, when allocating resources, attention should be paid to adopting a flexible allocation strategy to adapt to changes in demand and resource environment, and using an optimization algorithm to optimize resource allocation.

3. A multi-task reasoning scheduling method according to claim 1, characterized in that: In S4, when performing task priority classification, it is necessary to consider urgency, importance, dependency, task duration, resource availability, workload balance, risk-benefit and relevance requirements. It is necessary to assign higher priority to tasks with higher urgency and importance, and determine the priority ranking by considering the prerequisite relationship between tasks. Priority is assigned based on the availability of different resource types required, and higher priority is assigned to tasks with shorter duration and tasks with smaller workload balance. For high-risk tasks, a risk-benefit analysis is required, and multiple related tasks are uniformly ranked according to the relevance of the tasks.

4. A multi-task reasoning scheduling method according to claim 1, characterized in that: In S6, when scheduling resources, resources of a type with a larger amount of availability are scheduled preferentially.

5. A multi-task reasoning scheduling method according to claim 1, characterized in that: Said S10, adaptive adjustment of priorities and resource allocation should ensure that key individuals receive adequate resource support.

6. A multi-task reasoning scheduling device, characterized in that: include: Distributed processor units for independent processing of multiple tasks; Dynamic task scheduling unit, used to create, allocate and schedule resources for tasks; Collaborative reasoning unit, used to share information and knowledge between processors and perform joint reasoning; Resource optimization unit, used to monitor and feedback the execution of task scheduling and optimize the allocation mechanism; The dynamic task scheduling unit is connected to the resource optimization unit, and the distributed processor unit is connected to the dynamic task scheduling unit and the collaborative reasoning unit.

7. A multi-task reasoning scheduling device according to claim 6, characterized in that: The dynamic task scheduling unit includes a task creation unit, a resource evaluation unit, a task sorting unit, a resource allocation unit and a resource scheduling module. The task creation is used to create tasks and set priority parameters of tasks. The resource evaluation is used to evaluate the resource ratio occupied by tasks and system available resources. Task sorting is used to prioritize multiple tasks according to the priority parameters of the tasks and the resource ratio occupied. Resource allocation is used to allocate resources to each character according to the priority classification situation. Resource scheduling is used to schedule different resources according to the resource allocation situation.

8. A multi-task reasoning scheduling device according to claim 6, characterized in that: The resource optimization unit includes a feedback module, an evaluation module and an adjustment unit. The feedback unit is used to monitor task execution and resource utilization and to feed back monitoring information. The evaluation module is used to evaluate and count task execution and resource utilization, execution efficiency and execution reliability. The adjustment unit is used to adjust the logic of the task scheduling unit.

9. A multi-task reasoning scheduling device according to claim 6, characterized in that: The collaborative reasoning unit includes a resource collaboration module, a data collaboration module, a mechanism collaboration module and an intelligent collaboration module. The resource collaboration module is used to synchronize information and globally schedule system resources. The data collaboration module is used to collect data of distributed processor units and share them synchronously. The mechanism collaboration unit is used to synchronize the task allocation mechanism of distributed processor units. The intelligent collaboration module uses a deep learning algorithm to perform closed-loop optimization of distributed processor units.

10. The multi-task reasoning scheduling device according to claim 6, characterized in that: The distributed processor unit includes a communication module, a storage module, a processing module and a control module. The communication module is used to realize information transmission and instruction issuance between processors. The storage module is used to store, back up and restore task data and ensure the processing consistency of multiple servers. The processing module is used for task analysis, task execution and task feedback. The control module is used to realize control of interrupt devices.

Images