基于DFT-MARTE模型的时序分析算法  

Timing analysis algorithm based on DFT-MARTE model

作　　者：徐嘉 周晴 杜家昊 王一华 XU Jia;ZHOU Qing;DU Jia-hao;WANG Yi-hua(Key Laboratory of Electronic Information Technology for Complex Aerospace Systems,National Space Science Center,Chinese Academy of Sciences,Beijing 101499,China;School of Computer and Control,University of Chinese Academy of Sciences,Beijing 101408,China)

机构地区：[1]中国科学院国家空间科学中心复杂航天系统电子信息技术重点实验室,北京101499 [2]中国科学院大学计算机与控制学院,北京101408

出　　处：《计算机工程与设计》2024年第1期120-129,共10页Computer Engineering and Design

基　　金：民用航天技术预先研究基金项目(B0204)。

摘　　要：针对航天嵌入式软件(aerospace embedded software,AES)时序需求复杂带来的时序需求定义不准确问题,提出一种基于MARTE(modeling and analysis of real-time and embedded systems)模型的数据流时序(data flow timing based on MARTE,DFT-MARTE)模型,设计基于该模型的处理点缓存计算算法、时序偏离概率检测算法和时序序列分析算法。处理点缓存计算算法动态更新缓存空间,使后续时序检测正常执行;时序偏离概率检测算法利用多线程并发模拟时序特性,检测需求中时序偏离问题;时序序列分析算法是基于梯度下降算法,拟合时序序列,指导用户优化需求。该模型相比传统数据流模型更适用航天嵌入式软件,利于后续开发和维护,具有极高的应用价值。For the problem of the inaccurate definition of timing requirements caused by the complex timing requirements of aerospace embedded software(AES),a data flow timing based on modeling and analysis of real-time and embedded systems(DFT-MARTE)model was proposed,and a buffer calculation algorithm for processing points,a probability detection algorithm of timing deviations and an analysis algorithm of timing sequences were designed based on this model.The buffer calculation algorithm of processing points was used to dynamically update the cache space to make the subsequent timing detection execute normally.The probability detection algorithm of timing deviations was used to detect the timing deviation in requirements using multithreading concurrent simulation timing characteristics.The analysis algorithm of timing sequences was based on the gradient descent algorithm to fit the time series and guide users to optimize their needs.Compared with conventional data-flow models,this model is better suited for aerospace embedded software,which is advantageous for subsequent development and maintenance with high application value.

关 键 词：数据流时序模型 数据流图 嵌入式软件 时序偏离检测 多线程 时序分析 梯度下降算法 

分 类 号：TP311.5[自动化与计算机技术—计算机软件与理论]