Analysis And Research On Real-time Scheduling Of Time Uncertain Task Flow In Cyber-physical Systems

The Cyber-physical system(CPS)has strict requirements for real-time performance.In practical applications,the CPS system has various reasons such as the time granularity mismatch between the various transmission modules and the asynchronous synchronization of the collection tasks of the distributed system,causes the occurrence time of the tasks in the system to be within an interval,and the specific occurrence time in the interval is uncertain,this brings uncertainty to the system scheduling task and increases scheduling difficulty.At present,there are few research results to solve the problem that the task flow time in the CPS is uncertain and the dynamic real-time scheduling of the task flow.Some only solve a certain aspect without comprehensively considering the processing of the task time interval and the dynamic real-time scheduling of task flow.In this thesis,the time uncertain task flow in CPS is demonstrated by theoretical analysis,method research,case analysis and experimental verification.There are mainly the following aspects:(1)Based on the time uncertain task flow in CPS,the PPA scheduling algorithm is used to determine the initial scheduling sequence of the task flow.First,the task flow with uncertain time occurrence is combined to determine all possible scheduled sequences.Then,the effective upper and lower bounds of the time interval of the task flow in each combination sequence are determined by the optimal pruning algorithm,thereby shortening the length of the task time interval.For the same task,the effective upper and lower bounds are different in different combined sequences.Finally,the cyclic recursive method is used to calculate the probability that each combination sequence of the task flow is scheduled by the system,and then the probability that each task is first scheduled is determined.The probability is sorted in descending order to obtain the initial scheduling sequence of the task flow.(2)Based on the time uncertain task flow in the CPS,compared with the method in(1),another DS-IFS scheduling algorithm is used to determine the initial scheduling sequence of the task flow.Combining evidence theory(D-S theory)with Intuitionistic Fuzzy Set(IFS),Establishing the relationship between Belief Function(BeF)?Plausibility Function(PlF)in D-S theory and?_S(i)??_S(i)in IFS,predict the maximum possible moment of the task in the time interval.Firstly,the precise trust degree at the corresponding time is determined by the probability distribution of each time in the task time interval,thereby determining BeF and PlF at each moment,and then determining?_S(i)and?_S(i)at each moment.Then,based on the IFS negative time reasoning,the?_S(i)and?_S(i)of the occurrence time are derived according to the task end time and the residence time of the system.Finally,the maximum possible occurrence time of the discriminant prediction task is constructed by?_S(i)and?_S(i),and the initial scheduling sequence of the task flow is obtained by chronologically sorting the maximum possible occurrence time of each task.(3)Dynamic real-time scheduling analysis of the task flow with uncertain time in CPS.Traditional dynamic real-time scheduling algorithms are mostly based on the task occurrence time is determined,and there is no uncertainty about the time when the mission arrives at the system.Therefore,this thesis adopts a dynamic real-time scheduling algorithm for time uncertain task flow in CPS.First,the initial scheduled sequence of the task flow is determined by the method in(1)or(2),and then the task flow is dynamically scheduled in real time.In order to ensure the accuracy of the system scheduling tasks and ensure the system benefits,a dynamic priority assignment strategy combining task value and task urgency is proposed.At the same time,the system jitter avoidance method is adopted,and the jitter avoidance factor is set as the judgment basis of the task preemption,and the task execution is frequently switched during the system scheduling task.The case analysis and experimental results show that the real-time scheduling method for time uncertain task flow in CPS can solve the problem of task time uncertainty in CPS,at the same time,it can ensure the accuracy of the CPS real-time system when scheduling tasks,and ensure the success rate of task execution.Increase the difficulty of task scheduling,with the combination of(1)and(3)or(2)and(3),the system scheduling accuracy can reach 80%and 75%respectively.