The WCET Analysis Method For Big Data Driven Cyber-Physical System

Cyber-Physical System(CPS)is the core technology of many strategies such as "Industry 4.0" and "Made in China 2025".It is a system that integrates computing,communication and control.The feedback loop,which contains sensing,computing,decision-making and feedback control,realizes the fusion of the information world and the physical world.CPS is widely used in industrial manufacturing,intelligent transportation and other industries and life.CPS is a kind of system with high real-time requirements.Worst-Case Execution Time(WCET)analysis is one of the important foundations and inputs of task scheduling that CPS relies on to ensure real-time performance.This puts forward requirements for task WCET analysis of CPS,while the existing researches about task WCET analysis of CPS is rare.Nowadays,in the era of Big Data,besides that CPS itself is oriented to massive data,the data produced by big data analysis is comprehensive,accurate,and has wide coverage,which has high application value and can be used to assist CPS task WCET analysis.Therefore,on the basis of reviewing,analyzing and summarizing the relevant research and literature of CPS,WCET and big data,this paper considers and explores the relationship among them.After that,this paper proposes and establishes a complete and comprehensive WCET analysis method for big data-driven CPS.This paper first analyzes and points out that big data knowledge can help CPS to correct and narrow the input space and state set needed for WCET analysis,so as to increase the efficiency and accuracy of WCET analysis.On this basis,the flow path and definition of CPS tasks are given.The WCET of CPS task is decomposed into three parts: information layer(middleware)program WCET,network layer maximum transmission time and physical layer worst-case interaction time.And,the inclusion items and routes of WCET analysis of CPS are given accordingly.Further,the information-code-level analysis and low-level analysis are studied in depth,and the related technologies and tools involved in WCET analysis are elaborated in detail: control flow graph(CFG),Front End,control flow analysis(CFA),processor behavior analysis and timing anomalies,and three types of WCET estimation computing technologies: path-based,IPET-based,structure-based method.Thecorresponding examples are given separately.Finally,this paper adopts Simulink as an experimental tool,and takes the industrial temperature control system as an example to verify the WCET analysis method and idea of the big data-driven CPS proposed in this paper.This includes detailed analysis of the experimental system,hierarchical design and modeling,simulation and experimental data analysis.The experimental results prove the effectiveness and correctness of the WCET analysis method and ideas for the big data-driven CPS proposed in this paper,and also show that the big data knowledge can help improve the accuracy of WCET analysis.