Transmission and computing resource joint scheduling method for guaranteeing end-to-end real-time performance of industrial wireless control system

Abstract

The invention discloses a transmission and computing resource joint scheduling method for guaranteeing end-to-end real-time performance of an industrial wireless control system. The method comprises the following steps: acquiring a period, end-to-end deadline, task execution duration and message routing information of each control application in the industrial wireless control system; the end-to-end deadline is decomposed by setting the deadline of a calculation task in each control application; executing the following heuristic transmission scheduling based on the time constraint; setting the absolute deadline of the calculation task as the release time of the corresponding control message, and generating a calculation resource scheduling scheme based on an EDF algorithm; constructing a demand constraint function of the calculation tasks, and executing schedulability test for the calculation tasks of all the control applications; if all the control applications are schedulable, outputting a scheduling scheme; otherwise, adjusting system resources and rescheduling. According to the method, the coupling degree between two types of scheduling designs is effectively reduced, and meanwhile, the algorithm can be suitable for a controller based on a TDMA protocol.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of work automation, and in particular to a transmission and computing resource joint scheduling method for guaranteeing end-to-end real-time of industrial wireless control system. BACKGROUND

[0002] Industrial wireless control system becomes one of the key technologies for realizing intelligent transformation in the era of Industry 4.0, and is widely used in steel, automobile, metallurgy, electronics and other industries. Industrial wireless control system uses wireless network to realize reliable data transmission and collaborative control between different sensors, controllers, actuators and other devices, which can monitor and respond quickly to the production process, enhance the flexibility and real-time of the system, and thus improve the production efficiency and control accuracy. In typical control applications, sensors collect data and package them into sensing messages, which are transmitted to controllers through wireless communication. Controllers perform computing tasks to generate control instructions, which are packaged into control messages and transmitted to actuators to control the production process. If the time from data collection to execution of control instructions exceeds the end-to-end deadline, it will seriously affect the safe and stable operation of the control system. Therefore, ensuring that all control applications meet the end-to-end deadline requirement is a key task of industrial wireless control system.

[0003] In industrial wireless control system, the end-to-end response time of control applications is determined by message transmission delay and computing task execution delay. To improve the predictability and reliability of response time, the system uses a variety of optimization mechanisms. In terms of message transmission, industrial wireless network standards such as WirelessHART, ISA100.11a and 6TiSCH usually use time division multiple access (TDMA) to effectively reduce the collision probability of different messages in the transmission process under multi-hop network architecture by pre-planning time slot and channel allocation, thereby effectively reducing transmission delay. In terms of computing task scheduling, industrial wireless control system usually uses real-time operating systems such as FreeRTOS, RT-Thread, Linux-RT to ensure the real-time of tasks. In addition, real-time operating systems usually use scheduling algorithms such as earliest deadline first (EDF) and support task real-time preemption mechanism to improve the determinism of task execution. However, in industrial wireless control system, communication delay and computing delay affect each other, and focusing on a certain delay alone may lead to communication scheduling failure or computing task completion on time, resulting in incorrect control instructions and threatening system safety.

[0004] The invention with the patent application number CN 202010486304.9 and the name "industrial wireless network deterministic scheduling method supporting data packet aggregation" designs a transmission scheduling method based on conflict relaxation priority and maximum aggregation packet quantity priority, which can meet the deterministic transmission requirements of data flow reaching the destination device within the deadline. However, this method does not consider the scheduling of computing resources, which can easily lead to the failure of sending control messages in the planned time slot during the actual operation of the industrial wireless control system, thereby making it difficult to meet the end-to-end deadline constraints of control applications. The invention with the patent application number CN 202211052799.X and the name "joint allocation method of computing and communication resources of industrial wireless network" designs a joint allocation method of computing and communication resources based on a double-layer dual deep neural network model, but does not consider the industrial wireless communication protocol, and cannot be applied to industrial wireless control systems. The invention with the patent application number CN

[0005] 202411063971.0 and the name "industrial wireless network transmission and computing resource allocation method under imperfect channel" proposes a multi-agent deep reinforcement learning-based task offloading, multi-channel access and computing resource allocation method, which can effectively reduce the end-to-end delay of tasks. However, this method assumes that computing resources are unlimited, ignoring the fact that processors will allocate CPU to different tasks according to scheduling rules during actual operation. This assumption makes the estimation of the end-to-end delay of the task optimistic. During actual system operation, this may cause control applications to miss the end-to-end deadline, affecting the operation of the industrial wireless control system.

[0006] Therefore, it is crucial to design a joint scheduling method of transmission and computing resources to meet the end-to-end deadline constraints of control applications. SUMMARY

[0007] In view of the above-mentioned defects of the prior art, the technical problem to be solved by the present invention is to solve the joint scheduling method of transmission and computing resources to meet the end-to-end deadline constraints of control applications.

[0008] Industrial wireless control systems mainly include field devices (sensors and actuators) connected to industrial processes, controllers running control algorithms, and wireless access points as bridges between field devices and controllers. Control applications are periodic and consist of sensing tasks on sensors, computing tasks on controllers, and execution tasks on actuators. To facilitate information transmission, sensing tasks and computing tasks generate sensing messages and control messages, which are routed through wireless access points to trigger subsequent tasks. To meet the end-to-end deadline constraints of control applications, transmission resources need to be allocated to sensing messages and control messages, and computing resources need to be allocated to computing tasks.

[0009] In the existing research, the joint scheduling method of transmission and computing resources is usually designed under the assumption that the communication bandwidth and computing processor resources are infinitely divisible. However, in the industrial scene, the wireless communication network usually adopts the TDMA protocol, which divides the time into equal length time slots (usually 10ms level) and pre-allocates time slots and channels. At the same time, the real-time operating system usually allocates CPU resources to different tasks in a time-sharing manner according to the real-time scheduling algorithm such as EDF. The discrete characteristics of this resource allocation will inevitably cause the waiting time delay of messages and computing tasks in the process of waiting for the corresponding resources, and this key factor is ignored in the existing research. This makes the existing scheduling algorithm overestimate the end-to-end delay of the control application, which may cause the control application to miss the end-to-end deadline when the system is actually running, affecting the running effect of the industrial wireless control system.

[0010] To achieve the above purpose, the present application provides a transmission and computing resource joint scheduling method for guaranteeing the end-to-end real-time performance of an industrial wireless control system, comprising:

[0011] (1) obtaining the period, end-to-end deadline, task execution time and message routing information of each control application in the industrial wireless control system;

[0012] (2) decomposing the end-to-end deadline into:

[0013] the message transmission time limit from the completion of the sensing task to the start of the computing task;

[0014] the message transmission time limit from the completion of the computing task to the start of the execution task;

[0015] (3) based on the time constraints, performing the following heuristic transmission scheduling:

[0016] for sensing messages, using a forward scheduling strategy to select the earliest available time slot and channel that satisfies the constraints hop by hop from the first hop;

[0017] for control messages, using a backward scheduling strategy to select the latest time slot and channel that satisfies the constraints hop by hop from the last hop;

[0018] (4) setting the absolute deadline of the computing task as the release time of the corresponding control message, and generating a computing resource scheduling scheme based on the EDF algorithm;

[0019] (5) constructing the demand constraint function of the computing task, and performing schedulability check for the computing tasks of all control applications;

[0020] (6) if all control applications are schedulable, output the scheduling scheme; otherwise, adjust the system resources and return to step (2) to reschedule.

[0021] The application provides a transmission and computing resource joint scheduling algorithm, which converts end-to-end deadline constraints of an application into independent constraints of transmission scheduling and computing resource scheduling by exploring appropriate computing task deadline values, effectively reduces the coupling degree between the two types of scheduling design, and makes the algorithm applicable to industrial wireless networks based on a TDMA protocol and controllers deploying real-time operating systems.

[0022] Specifically, in terms of transmission scheduling, the algorithm can quickly allocate appropriate time slots and channel resources for data packets to meet the deterministic demand of industrial wireless networks on data transmission; in terms of task scheduling, the optimal task scheduling algorithm EDF is used to generate a computing resource scheduling scheme online, and the scheme can be directly applied to a real-time operating system.

[0023] Further, in the step (3), the transmission scheduling needs to meet the following constraint conditions:

[0024] The selected time slot of the current hop is later than the completion time slot of the previous hop or earlier than the start time slot of the next hop;

[0025] The sending node and the receiving node in the current time slot do not participate in other message transmission;

[0026] The selected channel is available in the current time slot and is not allocated to other messages;

[0027] The selected path hop number is consistent with the corresponding route of the message.

[0028] Further, the EDF scheduling algorithm selects the absolute deadline earliest one from all ready computing tasks for execution at each time point, and guarantees that the task completion time does not exceed the set deadline.

[0029] First, the transmission scheduling algorithm is designed according to the following rules: the control application is sorted according to the increasing order of the deadline, and the sensing message and control message are scheduled one by one. For the sensing message, time slots and channels are arranged hop by hop from the first hop, and the time slot and channel that meet all transmission constraints as early as possible are selected for transmission until the last hop is completed. The transmission constraints here include: the timing constraints of the sensing task and the sensing message, the time slot order constraints of the hop and the previous hop, the radio frequency constraints and the channel number constraints of the data packet sending and receiving nodes of the hop. For the control message, time slots and channels are arranged hop by hop from the last hop, and the time slot and channel that meet all transmission constraints as late as possible are selected for transmission until the first hop is completed. The transmission constraints here include: the timing constraints of the control message and the control task, the time slot order constraints of the hop and the next hop, the radio frequency constraints and the channel number constraints of the data packet sending and receiving nodes of the hop. This heuristic transmission scheduling algorithm relieves the transmission resource contention of the wireless access point by prioritizing the transmission of sensing messages and delaying the transmission of control messages as much as possible, and at the same time, it frees up more time for task execution, greatly improving the satisfaction rate of the end-to-end deadline of the control application.

[0030] Then, the absolute deadline of all computing tasks is set as the release time of the corresponding control message, and the EDF algorithm is directly configured and used in the real-time operating system to generate the computing schedule. Specifically, the real-time operating system will select the job with the earliest absolute deadline from a plurality of computing jobs at each time and execute it.

[0031] Further, the schedulability check includes:

[0032] A set of release times and absolute deadlines of all computing task instances within a hyper period is constructed;

[0033] A set of candidate time intervals is constructed, and each interval in the set has a start time and an end time selected from the set;

[0034] The cumulative execution demand of the tasks in each interval is calculated;

[0035] It is judged whether the demand of each interval exceeds the length of the interval.

[0036] In designing the joint scheduling method of transmission and computing resources, the existing research usually adopts two strategies. One is to construct a complex constraint problem according to the channel, time slot and other constraint conditions in the industrial wireless network. Most of such problems belong to NP-hard problems, and accurate solutions cannot be obtained in polynomial time. The second is to use a heuristic algorithm to simulate the computing and communication scheduling process of the entire system to ensure that all control applications meet the end-to-end deadline constraint.

[0037] However, the wireless network status in industrial environment is complex and changeable, which requires the generation of a computation and communication scheduling scheme in a short time to ensure the end-to-end deadline of control applications. Both of the above two strategies have the problem of long running time, and it is difficult to achieve a good balance between solution quality and real-time performance.

[0038] To quickly generate a computation and communication scheduling scheme that can ensure the end-to-end deadline of control applications, we further innovatively propose a schedulability test algorithm based on the proposed heuristic transmission and computation resource joint scheduling algorithm. This algorithm can quickly check whether the end-to-end deadline constraints of all control applications are met. At the same time, the schedulability test results can help system administrators reasonably configure transmission resources, such as increasing wireless access points, to improve the satisfaction rate of the end-to-end deadline of control applications.

[0039] Since the transmission scheduling has met the transmission constraints, the end-to-end deadline constraint test of all control applications can be converted into the EDF schedulability test of computation tasks, that is, to determine whether all computation jobs released by the computation task after receiving the sensor message can be completed before its absolute deadline.

[0040] First, the demand constraint function dbf(τ i,c ,t1,t2) is used to model the computation task τ i,c The maximum execution demand upper bound in the time interval [t1, t2] is the sum of the execution time of all computation jobs released after t1 and arriving at the absolute deadline before t2. Then, it is checked whether the sum of the demand constraint functions of the computation task set Γ in any time interval [t s ,t f ] satisfies the schedulability, that is, If all time intervals satisfy, it means that the transmission and computation scheduling can meet the end-to-end deadline constraints of all control applications, and the entire system can be started; if there is a time interval that does not satisfy, it means that the transmission or computation resource is insufficient, and engineers can increase the resource and run the heuristic joint scheduling algorithm again until the algorithm ends.

[0041] At the same time, to reduce the number of time intervals [t s ,t f ] to be checked and improve the speed of the algorithm to adapt to the rapidly changing wireless network status, we propose the following iteration method of t s and t f :

[0042] (1) t s ,t f only takes the release time or absolute deadline of a certain computation job, because dbf(τ i,c ,t s ,t f) is a step function, only in t s , f The release time or absolute deadline of a certain computing job is changed;

[0043] (2) If it is known that for a certain time interval [t s ,t f ] dbf (Γ, t s ,t f ) ≤ t f -t s , t f can be newly valued as less than the maximum release time or absolute deadline of a computing job dbf (Γ, t s ,t f )+t s , t s can be newly valued as greater than the minimum value t f -dbf (Γ, t s ,t f ).

[0044] This is because dbf (τ i,c ,t s ,t f ) is monotonically non-decreasing with t f , so it can be deduced that for any dbf (Γ, t s ,t' f ) ≤ dbf (Γ, t s ,t f ) ≤ t' f -t s . Similarly, it can be deduced that dbf (Γ, t' s ,t f ) ≤ t f -t' s . Thus, the number of time intervals [t s ,t f ] to be checked can be greatly reduced.

[0045] The schedulability test algorithm proposed by the application can quickly check whether the end-to-end deadline constraints of all control applications are met. Compared with the traditional method, the checking time is greatly shortened, the efficiency of system scheduling decision is improved, and unnecessary waste of computing resources is reduced.

[0046] In addition, in the case of variable wireless network in an industrial environment, the algorithm can reasonably allocate computing resources and transmission resources according to actual needs based on the fast inspection result, avoid the situation of over-allocation or under-allocation of resources, ensure the continuous normal operation of the system, and enhance the adaptability and robustness of the system to complex environments.

[0047] Further, to reduce the amount of calculation, only the task release time and the absolute deadline are reserved as the interval start and end boundaries to avoid redundant combinations.

[0048] Further, the method is applicable to an industrial wireless network using a TDMA protocol, which transmits data through statically pre-allocated time slots and channels.

[0049] Further, the method is applicable to an industrial controller environment deploying a real-time operating system, which supports task preemption and dynamic priority scheduling.

[0050] Further, the scheduling scheme is constructed with a control application super cycle as the basic time unit, and the super cycle is the least common multiple of all control application cycles.

[0051] The transmission and computing resource joint scheduling algorithm proposed in the application can efficiently generate a scheduling table that meets the deadlines of all control applications, and is applicable to various types of industrial embedded devices equipped with real-time operating systems and industrial wireless networks complying with the TDMA communication protocol.

[0052] The algorithm prioritizes the transmission of perception messages as early as possible and delays the transmission of control messages as much as possible, maximizes the time left for the transmission scheduling of the access point, and reserves more time for the execution of computing tasks, thereby improving the satisfaction rate of the end-to-end deadline of the control application.

[0053] Further, if the schedulability check fails, the method further comprises:

[0054] Determining the bottleneck resource type based on the check feedback;

[0055] Dynamically expanding the number of computing nodes or wireless access points;

[0056] Re-executing the joint scheduling algorithm after updating the system configuration.

[0057] Further, the method is embedded in an industrial wireless network scheduling manager module, supports online deployment, scheduling update and abnormal fast feedback of control applications, and is applicable to industrial automation scenarios such as metallurgy, electronics and automobile manufacturing with high real-time requirements.

[0058] Technical effects

[0059] The application provides a transmission and computing resource joint scheduling method for guaranteeing end-to-end real-time performance of an industrial wireless control system, which can be divided into two parts of a heuristic joint scheduling algorithm and a schedulability checking algorithm.

[0060] In terms of technical advantages, the NP-hard problem solving dilemma in the traditional method is innovatively avoided, and the computational complexity is greatly reduced. When facing the complex and changeable wireless network conditions in the industrial environment, the scheduling strategy can be quickly generated, and the schedulability can be quickly and accurately judged, thereby providing key technical support for guaranteeing the stable operation of the system.

[0061] In terms of performance indicators, the execution time of the scheduling algorithm can be effectively shortened, and the real-time performance of the system is significantly improved. At the same time, the end-to-end deadline constraints of all control applications can be quickly checked, the checking accuracy is extremely high, the risk of misjudgment is effectively avoided, and the safety and stability of the system are improved.

[0062] In terms of production implementation, the algorithm implementation process is simple, can be perfectly combined with the existing real-time embedded system and industrial wireless network protocol, does not need to make large-scale modification to the existing industrial system, and can be easily integrated into various industrial scenes, so that the application scope is wide. At the same time, in actual production, the scheduling hidden danger can be perceived in time, and the production link can be guided to make optimization and adjustment, thereby helping to improve the production efficiency and quality of the industrial wireless control system.

[0063] The concept, specific structure and generated technical effects of the application will be further described below with reference to the drawings, so as to fully understand the purpose, characteristics and effects of the application. BRIEF DESCRIPTION OF DRAWINGS

[0064] Figure 1 is a schematic diagram of an industrial wireless control system of a preferred embodiment of the application;

[0065] Figure 2 is a schematic diagram of a joint scheduling method of a preferred embodiment of the application;

[0066] Figure 3 is an algorithm flowchart of the constructed heuristic joint scheduling algorithm of a preferred embodiment of the application;

[0067] Figure 4 is an algorithm flowchart of the constructed schedulability checking algorithm of a preferred embodiment of the application. DETAILED DESCRIPTION

[0068] The technical content of the present application will be described more clearly and conveniently with reference to the accompanying drawings of the specification, and the present application can be embodied in many different forms, and the scope of protection of the present application is not limited to the embodiments described herein.

[0069] In the drawings, components of the same structure are denoted by the same reference numerals, and components having similar structures or functions are denoted by similar reference numerals. The size and thickness of each component shown in the drawings are arbitrarily shown, and the present application is not limited to the size and thickness of each component. In order to make the drawing clearer, the thickness of some components is appropriately exaggerated in some places in the drawing.

[0070] The present application provides a transmission and computing resource joint scheduling method for guaranteeing the end-to-end real-time performance of an industrial wireless control system, which can be divided into two parts: a heuristic joint scheduling algorithm and a schedulability checking algorithm. The heuristic joint scheduling algorithm does not rely on constructing a complex and difficult-to-solve constraint problem, but directly allocates transmission and computing resources reasonably according to the actual situation of channels, time slots, etc. in the industrial wireless network; the schedulability checking algorithm can quickly check whether the end-to-end deadline constraints of all control applications are met, ensuring the stable and safe operation of the entire industrial system.

[0071] As shown in Figure 1 , the industrial wireless control system mainly includes the following parts:

[0072] (1) field devices (sensors and actuators) connected to the industrial process;

[0073] (2) controllers running control algorithms;

[0074] (3) gateways as a bridge between field devices and controllers;

[0075] (4) multiple access points providing redundant communication paths that can directly connect field devices and gateways or indirectly connect through relay nodes;

[0076] (5) a network manager responsible for the management of the entire network;

[0077] (6) a host computer for remote monitoring and system configuration. All field devices, access points and relay nodes implement time synchronization, are equipped with half-duplex wireless transceivers, and comply with the TDMA protocol. Time is divided into fixed-length time slots, and each time slot is sufficient to complete a data packet transmission and its acknowledgement. Before the industrial wireless control system is started, the network manager will evaluate the channel status and list the channels with poor status in the blacklist to prohibit any device from using them.

[0078] The industrial wireless control system runs M control applications, each of which is composed of a sensing task on a sensor, a computing task on a controller, and an execution task on an actuator. Among them, the sensing task and the computing task trigger the subsequent tasks through sensing messages and control messages. In addition, each control application has a period, an end-to-end deadline, a route of sensing messages and control messages, a maximum execution time of sensing tasks, computing tasks and execution tasks. In order to meet the end-to-end deadline constraints of all control applications, the scheduling of all sensing messages, control messages, sensing tasks, computing tasks and execution tasks needs to be arranged within a super period H, where the super period H takes the minimum common multiple of the periods of all control applications.

[0079] As shown in Figure 2 The transmission and computing resource joint scheduling method for the industrial wireless control system described in this embodiment includes a heuristic joint scheduling algorithm and a schedulability checking algorithm. The heuristic joint scheduling algorithm arranges all control applications in order from small to large deadline, then allocates transmission resources for control applications one by one according to the actual situation of channels, time slots and the like in the industrial wireless network, and directly allocates computing resources based on the EDF algorithm; Then the schedulability checking algorithm can be used to quickly check whether the generated computing resource scheduling and transmission resource scheduling can meet the execution order constraints of sensing messages, computing tasks and control messages. When all control applications meet these constraints, return the scheduling result, and the algorithm ends. The entire system can be started, and when there is a control application that does not meet the constraints, it means that the system resources are insufficient, and the engineer can increase the system resources and run the heuristic joint scheduling algorithm again until the algorithm ends.

[0080] As shown in Figure 3 A heuristic joint scheduling algorithm is provided, and the steps are as follows:

[0081] Step 1: Obtain control application parameters and industrial wireless control system parameters, including period, deadline, route of sensing messages and control messages, maximum execution time of sensing tasks, computing tasks and execution tasks, time slot length, number of available channels;

[0082] Step 2: Sort the control applications in order from short to long deadline, and initialize the control application sequence number i = 1 currently considered;

[0083] Step 3: If there are still unscheduled control applications, the control application sequence number is incremented. When i < M, go to step 4; When i = M, generate the computing task scheduling of all control applications according to the EDF algorithm, and the algorithm ends.

[0084] Step 4: Assume

[0085] The sensing messages of this control application need s iThe jump data needs to be transmitted from the sensor to the controller. i The hop is transmitted from the controller to the actuator. Initialize the number of hops for the currently considered sensor message j = 1.

[0086] Step 5: If there are still unscheduled sensor message hop counts, increment the sensor message hop count. When j i If yes, proceed to step 6; otherwise, proceed to step 7.

[0087] Step 6: Find the earliest time slot t that satisfies the following scheduling constraints: (1) later than the completion time of the sensing task; (2) if j>1, later than the time slot allocated for the (j-1)th hop; (3) earlier than the end-to-end deadline of the control application; (4) within this time slot, the packet sending node and packet receiving node of this hop have not yet participated in the transmission of other messages; (5) within this time slot, there is an unallocated channel. If there is no time slot t that satisfies all the conditions, return the result of unschedulable, and the algorithm ends.

[0088] Step 7: Initialize the current control message hop count k = a i .

[0089] Step 8: If there are still unscheduled control message hops, decrease the control message hop count forward. If k > 1, proceed to step 9; otherwise, proceed to step 3.

[0090] Step 9: Find the latest time slot t that satisfies the following scheduling constraints: (1) earlier than the start time of the task execution; (2) if k i (2) It must be earlier than the time slot allocated for the (k+1)th hop; (3) It must be later than the completion time of the sensing task; (4) Within this time slot, the data packet sending node and data packet receiving node of this hop have not yet participated in the transmission of other messages; (5) Within this time slot, there is an unallocated channel. If there is no time slot t that satisfies all the conditions, return the result of unschedulable, and the algorithm ends.

[0091] like Figure 4 As shown, a schedulability verification algorithm is provided, with the following steps:

[0092] Step 1: Determine the release time and absolute deadline of all instances of all computing tasks within a supercycle. These time points constitute the set D of time points to be checked.

[0093] Step 2: Assign the smallest value in D to t. s Assign the maximum value in D to t. max .

[0094] Step 3: If t s Less than supercycle t f Initialize to t max ​​Otherwise, return "schedulable" directly, and the algorithm ends.

[0095] Step 4: If t s < t f , calculate the total execution demand of all computing jobs whose release time and absolute deadline are in the time interval [t s , t f ], denoted as dbf(t s , t f ), and then go to Step 5; otherwise, update t s to the minimum time value in set D that satisfies the condition t f - dbf(t s , t max ), and then go to Step 3.

[0096] Step 5: If dbf(t s , t f ) is greater than t f - t s , it indicates that the time resource required by the control application in this time interval exceeds the total length of the interval, and the algorithm ends with returning "unschedulable"; otherwise, update t f to the maximum time value in set D that is less than or equal to t s + dbf(t s , t f ), and then go to Step 4.

[0097] The preferred embodiments of the present application are described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and variations without departing from the concept of the present application. Therefore, the technical solutions obtained by logical analysis, reasoning or limited experiments based on the prior art according to the concept of the present application should also be within the scope of protection defined by the claims.

Claims

1. A method for jointly scheduling transmission and computing resources to ensure end-to-end real-time performance of an industrial wireless control system, characterized in that, The method comprises the following steps: (1) obtaining the period, end-to-end deadline, task execution duration and message routing information of each control application in the industrial wireless control system; (2) decomposing the end-to-end deadline into the following two parts by setting the computation task deadline in each control application: message transmission time limit from the completion of the sensing task to the start of the computation task; message transmission time limit from the completion of the computation task to the start of the execution task; (3) performing the following heuristic transmission scheduling based on the time constraints: for sensing messages, using a forward scheduling strategy to select the earliest available time slot and channel that meet the constraints hop by hop from the first hop; for control messages, using a backward scheduling strategy to select the latest time slot and channel that meet the constraints hop by hop from the last hop; (4) setting the absolute deadline of the computation task as the release time of the corresponding control message, and generating a computation resource scheduling scheme based on the EDF algorithm; (5) constructing a demand constraint function of the computation task, and performing schedulability verification for the computation tasks of all control applications; (6) if all control applications are schedulable, outputting the scheduling scheme; otherwise, adjusting the system resources and returning to step (2) to reschedule.

2. The method of claim 1, wherein, In step (3), the transmission scheduling needs to meet the following constraint conditions: the selected time slot of the current hop is later than the completion time slot of the previous hop or earlier than the start time slot of the next hop; the sending node and the receiving node in the current time slot do not participate in other message transmission; the selected channel is available in the current time slot and is not allocated to other messages; the selected path hop number is consistent with the corresponding routing of the message.

3. The method of claim 1, wherein, The EDF scheduling algorithm selects the computation task with the earliest absolute deadline from all ready computation tasks at each time point for execution, so as to ensure that the task completion time does not exceed the set deadline.

4. The method of claim 1, wherein, The schedulability verification comprises the following steps: constructing a set of release times and absolute deadlines of all computation task instances within a super period; constructing a candidate time interval pair, and selecting the start and end points of each interval from the set; calculating the cumulative execution demand of the tasks in each interval; judging whether the demand of each interval exceeds the interval length.

5. The method of claim 4, wherein, In order to reduce the calculation amount, only the release time and the absolute deadline of the task are reserved as the interval start and end boundaries, and redundant combinations are avoided.

6. The method of claim 1, wherein, The method is suitable for an industrial wireless network using a TDMA protocol, and the network transmits data through statically pre-allocated time slots and channels.

7. The method of claim 1, wherein, The method is suitable for an industrial controller environment deploying a real-time operating system, and the real-time operating system supports task preemption and dynamic priority scheduling.

8. The method of claim 1, wherein, The scheduling scheme is constructed in a control application super period as a basic time unit, and the super period is the least common multiple of the periods of all control applications.

9. The method of claim 1, wherein, If the schedulability verification fails, the method further comprises the following steps: determining the bottleneck resource type based on the verification feedback; dynamically expanding the number of computation nodes or wireless access points; updating the system configuration and re-executing the joint scheduling algorithm.

10. The method of claim 1, wherein, The method is embedded in an industrial wireless network scheduling manager module, supports online deployment, scheduling update and abnormal fast feedback of control applications, and is suitable for metallurgy, electronics, automobile manufacturing and other industrial automation scenarios with high real-time requirements.

Images