Embedded multi-core operating system scheduling method and device and storage medium

Abstract

The invention discloses an embedded multi-core operating system scheduling method and device and a storage medium, and particularly relates to the technical field of operating system scheduling. Comprising the steps of S01, processor core numbering, S02, system execution task data collection, S03, task suitability evaluation, S04, processor core energy consumption monitoring, S05, task intelligent scheduling and S06, system execution task fault detection. The method can effectively manage a plurality of processor cores, ensure efficient execution of tasks, trigger energy consumption early warning, execute task scheduling and improve the energy utilization efficiency of the system by monitoring core operation energy consumption evaluation indexes, and improve the energy utilization efficiency of the system by quantifying the relationship between fault duration and energy consumption indexes. Fault detection and task migration of a processor core are realized, and the stability of an industrial operating system is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of operating system scheduling, more particularly, the present application relates to an embedded multi-core operating system scheduling method and device and storage medium. BACKGROUND

[0002] With the wide application of embedded systems in aerospace, industrial control, smart home and other fields, the performance requirements of embedded multi-core operating systems are becoming higher and higher. Industrial automation systems integrate more and more devices and functions, and the pros and cons of task scheduling strategy directly affect the overall performance, real-time performance and stability of the system.

[0003] CN114168352A discloses a multi-core scheduling method and device for tasks, electronic equipment and storage medium, which selects specific tasks in the running task queue and the ready task queue, and preempts based on priority to avoid the problem of reduced multi-core processing performance in the task preemption process, CN115809132A discloses a lightweight and scalable multi-core embedded operating system scheduler design method, which designs a scheduler containing a scheduler main framework, a scheduler state machine and a scheduler queue from the aspects of architecture, scalability and performance, proposes an abstract scheduler model and an extensible plug-in scheduler framework, and optimizes the ready queue management.

[0004] However, in actual use, there are still some shortcomings, such as the existing embedded system in the multi-device simultaneous work scene still has room for improvement in the matching response of task scheduling and device resources, resulting in insufficient overall stability and unreasonable resource allocation during system operation. Traditional task scheduling methods often only focus on the execution efficiency of tasks and do not consider the energy consumption problem in different task execution processes, lack of scheduling methods based on energy consumption perception, and if high-energy consumption devices execute low-load tasks for a long time, the device load is unbalanced, affecting the overall production efficiency. SUMMARY

[0005] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present application provide an embedded multi-core operating system scheduling method, device and storage medium to solve the problems raised in the background art.

[0006] To achieve the above-mentioned purpose, the present application provides the following technical scheme: an embedded multi-core operating system scheduling method, comprising the following steps: Step S01: processor core number: sequentially number the plurality of processor cores of the target industrial operating system as 1, 2,... i,... n.

[0007] Step S02: System performs task data collection: for collecting task-core matching data of each processor core of the target industrial operation system, the step S02: system performs task data collection includes task resource demand data collection sub-step and core energy consumption data collection sub-step, and the task-core matching data includes task resource demand data and core energy consumption data.

[0008] Step S03: Task adaptability evaluation: for receiving the task-core matching data transmitted by the system performing task data collection step, calculating the adaptability score of each processor core of the target industrial operation system according to the task resource demand data collected by the task resource demand data collection sub-step, and selecting the processor core with low adaptability score to allocate the task.

[0009] Step S04: Processor core energy consumption monitoring: for receiving the task-core matching data transmitted by the system performing task data collection step, calculating the core running energy consumption evaluation index of each processor core of the target industrial operation system according to the core energy consumption data collected by the core energy consumption data collection sub-step.

[0010] Step S05: Task intelligent scheduling: obtaining the core running energy consumption evaluation index of each processor core of the target industrial operation system, comparing with the preset core running energy consumption evaluation index, and judging whether task scheduling is needed.

[0011] Step S06: System performs task fault detection: based on the fault duration of each processor core of the target industrial operation system, combining the core running energy consumption evaluation index of each processor core, calculating the fault feedback coefficient of each processor core of the target industrial operation system, and performing fault judgment.

[0012] Preferably, the step S02: system performs task data collection is specifically: Task resource demand data collection sub-step: collecting the task queue length, the maximum allowed task queue length, the CPU usage rate, and the task urgency level (0-10 level) of each processor core of the target industrial operation system, respectively marked as 、 、 、 ; Core energy consumption data collection sub-step: collecting the power consumption and temperature of each processor core of the target industrial operation system, respectively marked as 、 .

[0013] Preferably, the step S03: task adaptability evaluation is specifically: S31: Calculate the adaptability score of each processor core through the task queue length of each processor core, the maximum allowed task queue length, CPU usage, and task urgency; S32: Obtain the adaptability score of each processor core, arrange them in descending order of adaptability score value, the higher the adaptability score, the more tasks running on the core, the busier the core, then allocate the new task to the idle core.

[0014] Preferably, the step S04: processor core energy consumption monitoring specifically includes: Obtain the power consumption of each processor core through a hardware sensor, obtain the temperature of each processor core through a thermistor, and calculate the core running energy consumption evaluation index of each processor core of the target industrial operating system.

[0015] Preferably, the step S05: task intelligent scheduling specifically includes: S51: Collect the core running energy consumption evaluation index of each processor core of the target industrial operating system every j time period; S52: Compare the core running energy consumption evaluation index of each processor core with the preset core running energy consumption evaluation index, if the core running energy consumption evaluation index of a certain processor core is greater than the preset core running energy consumption evaluation index, it indicates that the processor core is a high-energy-consumption core, then trigger the energy consumption warning and enter the task scheduling process, at this time the industrial operating system obtains the task completion degree of the tasks executed by each processor core, otherwise it indicates that the processor core is an adaptive and low-energy-consumption core, and does not enter the task scheduling; S53: Evaluate the schedulability of the processor core based on the task completion degree, the higher the task completion degree of the processor core, the lower the migration cost of the processor core, the task scheduling is preferentially executed, compare the task completion degree of the tasks executed by each processor core with the set task completion degree threshold, if the task completion degree of the tasks executed by a certain processor core is less than or equal to the set task completion degree threshold, it indicates that the migration cost of the processor core is high, then retain the tasks executed by the processor core, otherwise, sort the processor cores whose task completion degree is greater than the set task completion degree threshold in ascending order of task completion degree; S54: Obtain all processor cores with high energy consumption in the target industrial operating system, select the optimal schedulable core for the tasks executed by the high-energy-consumption core through the industrial operating system, and distribute the target task to the schedulable target processor core.

[0016] Preferably, the step S06: system task fault detection specifically includes: S61: Obtain the fault duration of each processor core of the target industrial operating system; S62: calculating the fault feedback coefficient of each processor core of the target industrial operation system through the fault duration of each processor core, the maximum fault duration, the minimum fault duration, and the core running energy consumption evaluation index; S63: obtaining the fault feedback coefficient of each processor core of the target industrial operation system, comparing the fault feedback coefficient with a preset fault feedback coefficient, if the fault feedback coefficient of a certain processor core is greater than the preset fault feedback coefficient, it indicates that the processor core has a fault, and the execution of the task is terminated immediately, a fault alarm is sent to the manager, and the executed task is allocated to a schedulable processor core, otherwise, it indicates that the task running of the processor core is normal.

[0017] Preferably, the embedded multi-core operation system scheduling device comprises a memory, a processor, and a machine executable program stored in the memory and running on the processor, and the processor implements the embedded multi-core operation system scheduling method when executing the machine executable program.

[0018] Preferably, the computer readable storage medium stores a computer program.

[0019] The technical effects and advantages of the present application are as follows: 1. The present application provides an embedded multi-core operation system scheduling method, device and storage medium, by collecting the task-core matching data of each processor core of the target industrial operation system, calculating the adaptability score of each processor core based on the task queue length, the maximum allowed task queue length, the CPU usage rate, and the task urgency of each processor core, arranging the adaptability score values from high to low, the higher the adaptability score, the more tasks running on the core, and the more busy the core is, then allocating new tasks to idle cores, dynamically adjusting the task allocation according to the processing capacity of the processor core, which can effectively manage multiple processor cores, ensure efficient task execution, monitor the power consumption and temperature of each processor core, calculate the core running energy consumption evaluation index of each processor core of the target industrial operation system, and the unbalanced task allocation makes the energy consumption of part of the core too high, high energy consumption is accompanied by high temperature, by monitoring the energy consumption state of the processor core, the problem of excessive energy consumption of the processor core in the industrial operation system can be effectively solved, and the task allocation can be reasonably adjusted. 2、The application provides a kind of embedded multi-core operating system scheduling method, device and storage medium, by every j Time interval collection target industrial operating system each processor core core running energy consumption evaluation index, with the comparison of preset core running energy consumption evaluation index, if the core running energy consumption evaluation index of certain processor core is greater than the preset core running energy consumption evaluation index, then indicate that the processor core is high-energy core, then trigger energy consumption early warning, enter task scheduling process, at this time industrial operating system obtains the task unfinished degree of each processor core executed task, the task unfinished degree of each processor core executed task is compared with the set task unfinished degree threshold, the processor core with the task unfinished degree less than or equal to the set task unfinished degree threshold is sorted in ascending order of task unfinished degree, executes task scheduling, and the task executed by high-energy core is selected by industrial operating system The optimal schedulable core and the target task is distributed to the schedulable target processor core, the uneven task distribution makes part of core energy consumption too high, by monitoring processor core energy consumption state and intelligently adjusting task distribution, realize the task balanced allocation of processor core, can effectively solve the problem of high energy consumption of processor core in industrial operating system, improve system energy utilization efficiency, further by the fault duration of each processor core, in combination with the core running energy consumption evaluation index of each processor core, the fault feedback coefficient of each processor core of target industrial operating system is calculated, fault determination is carried out, the relationship between quantitative fault duration and energy consumption index is constructed Fault feedback model, realize the fault detection and task migration of processor core, improve the stability of industrial operating system. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 It is the flowchart of the application of a kind of embedded multi-core operating system scheduling method. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0022] Please refer to Figure 1 The application provides a kind of embedded multi-core operating system scheduling method, including the following steps: The step S01: processor core number: the multiple processor cores of target industrial operating system are numbered as 1, 2,... i,... n in turn; The step S02: the system performs task data collection: used for collecting task-core matching data of each processor core of the target industrial operation system, the step S02: the system performs task data collection includes a task resource requirement data collection substep and a core energy consumption data collection substep, and the task-core matching data includes task resource requirement data and core energy consumption data; In a possible design, the step S02: the system performs task data collection specifically includes: The task resource requirement data collection substep: collecting task queue length, allowed maximum task queue length, CPU usage, and task urgency level (0-10 levels) of each processor core of the target industrial operation system, and marking them as , , , wherein i=1, 2,...n, i represents the number of the ith processor core; The core energy consumption data collection substep: collecting power consumption and temperature of each processor core of the target industrial operation system, and marking them as , .

[0023] The step S03: task adaptability evaluation: used for receiving the task-core matching data transmitted by the system performing the task data collection step, calculating the adaptability score of each processor core of the target industrial operation system according to the task resource requirement data collected in the task resource requirement data collection substep, and selecting a processor core with a low adaptability score to allocate a task; dynamically adjusting the task allocation according to the processing capacity of the processor core, which can effectively manage multiple processor cores and ensure efficient task execution.

[0024] In a possible design, the step S03: task adaptability evaluation specifically includes: S31: calculating the adaptability score of each processor core by using the task queue length, allowed maximum task queue length, CPU usage, and task urgency level of each processor core; S32: obtaining the adaptability score of each processor core, arranging the adaptability scores in descending order according to the adaptability score values, and the higher the adaptability score, the more tasks running on the core, and the busier the core, so that a new task is allocated to an idle core; In this embodiment, it needs to be specifically explained that the calculation formula of the adaptability score is as follows: wherein, represents the adaptability score of the ith processor core, represents the task queue length of the ith processor core, a maximum task queue length allowed by the i-th processor core, a CPU usage rate of the i-th processor core, a task urgency of the i-th processor core, , , weight coefficients of the task queue length, the CPU usage rate and the task urgency, respectively, and = 1.

[0025] The step S04: processor core energy consumption monitoring is configured to receive the task-core matching data transmitted by the system execution task data collection step, calculate the core running energy consumption evaluation index of each processor core of the target industrial operation system according to the core energy consumption data collected by the core energy consumption data collection sub-step; the unbalanced task allocation causes the energy consumption of some cores to be too high, and high energy consumption is accompanied by high temperature; by monitoring the energy consumption state of the processor core, the problem of too high energy consumption of the processor core in the industrial operation system can be effectively solved, and the task allocation can be reasonably adjusted.

[0026] In a possible design, the step S04: processor core energy consumption monitoring specifically includes: S41: acquiring the power consumption of each processor core through a hardware sensor, acquiring the temperature of each processor core through a thermistor, and calculating the core running energy consumption evaluation index of each processor core of the target industrial operation system.

[0027] In this embodiment, it is specifically pointed out that the calculation formula of the core running energy consumption evaluation index is: wherein, the core running energy consumption evaluation index of the i-th processor core, the power consumption of the i-th processor core, the temperature of the i-th processor core, the maximum power consumption of the processor core, the maximum temperature of the processor core.

[0028] The step S05: task intelligent scheduling is configured to acquire the core running energy consumption evaluation index of each processor core of the target industrial operation system, compare the core running energy consumption evaluation index with a preset core running energy consumption evaluation index, and determine whether task scheduling is needed; the unbalanced task allocation causes the energy consumption of some cores to be too high; by monitoring the energy consumption state of the processor core and intelligently adjusting the task allocation, the task allocation of the processor core can be balanced, the problem of too high energy consumption of the processor core in the industrial operation system can be effectively solved, and the energy utilization efficiency of the system can be improved.​​

[0029] In a possible design, the step S05: task intelligent scheduling specifically includes the following steps. S51: collect the core running energy consumption evaluation index of each processor core of the target industrial operation system every j time periods; S52: compare the core running energy consumption evaluation index of each processor core with a preset core running energy consumption evaluation index, if the core running energy consumption evaluation index of a certain processor core is greater than the preset core running energy consumption evaluation index, it indicates that the processor core is a high-energy-consumption core, and an energy consumption warning is triggered, and the task scheduling process is entered, at this time, the industrial operation system obtains the task completion degree of the task executed by each processor core, otherwise, it indicates that the processor core is a low-energy-consumption core with optimal adaptability, and task scheduling is not entered; S53: evaluate the schedulability of the processor core based on the task completion degree, the higher the task completion degree of the processor core, the lower the migration cost of the processor core, and the task scheduling is preferentially executed, compare the task completion degree of the task executed by each processor core with a set task completion degree threshold, if the task completion degree of the task executed by a certain processor core is less than or equal to the set task completion degree threshold, it indicates that the migration cost of the processor core is high, and the task executed by the processor core is retained, otherwise, the task scheduling is performed on the unfinished task, and the processor cores with a task completion degree greater than the set task completion degree threshold are sorted in ascending order of the task completion degree; S54: obtain all processor cores with high-energy-consumption cores in the target industrial operation system, distribute the task executed by the high-energy-consumption core to the target processor core through the industrial operation system to select the optimal schedulable core and distribute the target task to the schedulable target processor core.

[0030] The step S06: system task fault detection specifically includes the following steps.

[0031] In a possible design, the step S06: system task fault detection specifically includes the following steps. S61: obtain the fault duration of each processor core of the target industrial operation system; S62: calculate the fault feedback coefficient of each processor core of the target industrial operation system based on the fault duration, the maximum fault duration, the minimum fault duration, and the core running energy consumption evaluation index of each processor core; S63: Obtain the fault feedback coefficient of each processor core of the target industrial operation system, compare it with the preset fault feedback coefficient, if the fault feedback coefficient of a processor core is greater than the preset fault feedback coefficient, it indicates that the processor core has a fault, immediately terminate the execution of the task, send a fault alarm to the manager, and assign the executed task to the schedulable processor core, otherwise, it indicates that the task running of the processor core is normal.

[0032] In the embodiment, it needs to be particularly pointed out that the calculation formula of the fault feedback coefficient of each processor core is: Among them, represents the fault feedback coefficient of the i th processor core, represents the core running energy consumption evaluation index of the i th processor core, represents the minimum value of the core running energy consumption evaluation index, represents the maximum value of the core running energy consumption evaluation index, represents the fault duration of the i th processor core, represents the maximum value of the fault duration, represents the minimum value of the fault duration.

[0033] The preset value and the threshold value are set for the purpose of comparison. The specific values of the preset value and the threshold value depend on the number of sample data and the base number set by the person skilled in the art for each group of sample data. As long as it does not affect the proportional relationship between the parameters and the quantized values.

[0034] The embedded multi-core operation system scheduling device comprises a memory, a processor, and a machine executable program stored on the memory and running on the processor, and the processor implements the embedded multi-core operation system scheduling method when executing the machine executable program.

[0035] The computer readable storage medium has a computer program stored thereon.

[0036] In the embodiment, it needs to be particularly pointed out that the application calculates the adaptability score of each processor core based on the task queue length, the maximum allowed task queue length, the CPU usage rate and the task urgency of each processor core by collecting the task-core matching data of each processor core of the target industrial operation system, arranges the adaptability score values from high to low, the higher the adaptability score, the more the task quantity running on the core, the busier the core, and then the new task is distributed to the idle core, the task distribution is dynamically adjusted according to the processing capacity of the processor core, the multiple processor cores can be effectively managed, the task efficient execution is ensured, the power consumption and temperature of each processor core are monitored, the core running energy consumption evaluation index of each processor core of the target industrial operation system is calculated, the unbalanced task distribution makes the energy consumption of part of the core too high, the high energy consumption is accompanied by high temperature, the problem of too high energy consumption of the processor core in the industrial operation system can be effectively solved by monitoring the energy consumption state of the processor core, and the task distribution is reasonably adjusted.

[0037] The application collects the core running energy consumption evaluation index of each processor core of the target industrial operation system every j time period, compares the core running energy consumption evaluation index with the preset core running energy consumption evaluation index, if the core running energy consumption evaluation index of a certain processor core is greater than the preset core running energy consumption evaluation index, it indicates that the processor core is a high energy consumption core, then the energy consumption early warning is triggered, and the task scheduling process is entered, at this time, the industrial operation system obtains the task unfinished degree of the task executed by each processor core, compares the task unfinished degree of the task executed by each processor core with the set task unfinished degree threshold, sorts the processor cores with the task unfinished degree less than or equal to the set task unfinished degree threshold in ascending order of the task unfinished degree, executes the task scheduling, and distributes the target task to the schedulable target processor core selected by the industrial operation system, the unbalanced task distribution makes the energy consumption of part of the core too high, the task balanced distribution of the processor core is realized by monitoring the energy consumption state of the processor core and intelligently adjusting the task distribution, the problem of too high energy consumption of the processor core in the industrial operation system can be effectively solved, the system energy utilization efficiency is improved, the fault feedback coefficient of each processor core of the target industrial operation system is calculated by combining the core running energy consumption evaluation index of each processor core and the fault duration of each processor core, the fault is judged, the relationship between the fault duration and the energy consumption index is quantified, the fault feedback model is constructed, the fault detection and task migration of the processor core are realized, and the stability of the industrial operation system is improved.

[0038] Finally: the above only describes the preferred embodiments of the application and is not used to limit the application, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the application shall be included in the protection scope of the application.

Claims

1. An embedded multi-core operating system scheduling method, characterized by, Comprise the following steps: Step S01: processor core number: the target industrial operating system is numbered as 1, 2,... i,... n in turn; Step S02: system execution task data collection: for collecting the task-core matching data of each processor core of the target industrial operating system, the step S02: system execution task data collection includes task resource demand data collection substep and core energy consumption data collection substep, the task-core matching data includes task resource demand data and core energy consumption data; Step S03: task adaptability evaluation: for receiving the task-core matching data transmitted by the system execution task data collection step, calculating the adaptability score of each processor core of the target industrial operating system according to the task resource demand data collected by the task resource demand data collection substep, and selecting the processor core with low adaptability score to allocate tasks; Step S04: processor core energy consumption monitoring: for receiving the task-core matching data transmitted by the system execution task data collection step, calculating the core running energy consumption evaluation index of each processor core of the target industrial operating system according to the core energy consumption data collected by the core energy consumption data collection substep; Step S05: task intelligent scheduling: obtaining the core running energy consumption evaluation index of each processor core of the target industrial operating system, comparing with the preset core running energy consumption evaluation index, and judging whether task scheduling is needed; Step S06: system execution task fault detection: based on the fault duration of each processor core of the target industrial operating system, combining the core running energy consumption evaluation index of each processor core, calculating the fault feedback coefficient of each processor core of the target industrial operating system, and performing fault judgment.

2. The method of claim 1, wherein: The step S02: system execution task data collection is specifically: Task resource demand data collection sub-step: collect the task queue length, the allowed maximum task queue length, the CPU usage rate, and the task urgency of each processor core of the target industrial operation system, respectively marked as , , , ; Core energy consumption data collection sub-step: collect the power consumption and temperature of each processor core of the target industrial operation system, respectively marked as , .

3. The method of claim 1, wherein: The step S03: task adaptability evaluation is specifically: S31: calculating the adaptability score of each processor core through the task queue length, the maximum allowed task queue length, CPU usage, and task urgency of each processor core; S32: obtaining the adaptability score of each processor core, arranging according to the adaptability score value from high to low, the higher the adaptability score, the more tasks running on the core, the busier the core, then the new task is allocated to the idle core.

4. The method of claim 1, wherein: The step S04: processor core energy consumption monitoring is specifically: Through hardware sensor, the power consumption of each processor core is obtained, and through thermistor, the temperature of each processor core is obtained, and the core running energy consumption evaluation index of each processor core of the target industrial operating system is calculated.

5. The method of claim 1, wherein: The step S05: task intelligent scheduling is specifically: S51: collecting the core running energy consumption evaluation index of each processor core of the target industrial operating system every j time period; S52: compare the core running energy consumption evaluation index of each processor core with the preset core running energy consumption evaluation index, if the core running energy consumption evaluation index of a certain processor core is greater than the preset core running energy consumption evaluation index, it indicates that the processor core is a high energy consumption core, then trigger energy consumption warning and enter the task scheduling process, at this time the industrial operating system obtains the task completion degree of the task executed by each processor core, otherwise, it indicates that the processor core is an adaptive and low energy consumption core, and does not enter task scheduling; S53: evaluate the schedulability of the processor core based on the task completion degree, the higher the task completion degree of the processor core, the lower the migration cost of the processor core, and the task scheduling is preferentially executed, compare the task completion degree of the task executed by each processor core with the set task completion degree threshold, if the task completion degree of the task executed by a certain processor core is less than or equal to the set task completion degree threshold, it indicates that the migration cost of the processor core is high, and the task executed by the processor core is retained, otherwise, the task scheduling is performed on the unfinished task, and the processor cores with the task completion degree greater than the set task completion degree threshold are sorted in ascending order of the task completion degree; S54: obtain all processor cores of the target industrial operating system in a high energy consumption core, and distribute the task executed by the high energy consumption core to the schedulable target processor core through the industrial operating system.

6. The method of claim 1, wherein: The step S06: the system executes task fault detection specifically comprises: S61: obtain the fault duration of each processor core of the target industrial operating system; S62: calculate the fault feedback coefficient of each processor core of the target industrial operating system through the fault duration, the maximum fault duration, the minimum fault duration and the core running energy consumption evaluation index of each processor core; S63: compare the fault feedback coefficient of each processor core of the target industrial operating system with the preset fault feedback coefficient, if the fault feedback coefficient of a certain processor core is greater than the preset fault feedback coefficient, it indicates that the processor core has a fault, immediately terminate the execution of the task and issue a fault alarm to the manager, and distribute the executed task to the schedulable processor core, otherwise, it indicates that the task running of the processor core is normal.

7. An embedded multi-core operating system scheduling apparatus, characterized by: The computer program product comprises a memory, a processor and a machine executable program stored on the memory and running on the processor, and the processor implements the embedded multi-core operating system scheduling method according to any one of claims 1-6 when executing the machine executable program.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to implement the embedded multi-core operating system scheduling method according to any one of claims 1-6.

Images