TY - JOUR
T1 - 面向轨交控制软件需求模型的量纲分析方法
AU - Wang, Shang
AU - Feng, Jin Cao
AU - Zhu, Jia Yi
AU - Huang, Yi Hao
AU - Zheng, Han Yue
AU - Xu, Xiang Rong
AU - Miao, Wei Kai
AU - Zhang, Xiang
AU - Pu, Ge Guang
N1 - Publisher Copyright:
© 2020, Science Press. All right reserved.
PY - 2020/11
Y1 - 2020/11
N2 - Embedded control software is the kernel component of today's control systems. For the safety-critical control systems, such as the railway system, aerospace control system and nuclear power control system, the function and safety of the control software embedded in them are directly related to the safety of people's lives and property. After long-term research and practice, both academia and industry realized that system requirements are the starting point of the software engineering life cycle and the fundamental way to ensure software quality. Therefore, how to ensure the correctness of software development process from the source has been an important research topic in the field of software engineering. However, in the software development of industry, the correctness of system requirements can only be guaranteed by manual review. Manual review of system requirements described in natural language is not only inefficient, but also unable to ensure the correctness of requirements. There is still a lack of a formal modeling analysis method to validate requirements. Requirements validation is an important engineering activity to ensure the quality of the formalized specifications constructed. As an important method of requirements validation, dimensional analysis can check out potential defects caused by wrong definition of dimension and reduce errors caused by improper use of dimensions. At present, the dimensional analysis in computer systems is mostly based on the code level and not applied in the requirements field, while some hidden errors in safety-critical embedded control systems are mostly caused by the wrong use of dimensions in requirements, which leads to software development errors from the source. Therefore, we propose a requirements modeling and dimension analysis method by according to the characteristics of the railway, which is based on formal engineering method. First, the requirements engineer converts the original requirements specifications into the requirements specifications described using the domain-specific requirements description language, then automatically extracts the requirements model using the tool. Finally, carries out the subsequent dimensional analysis on the basis of the requirements model. This work combines with the basic principle of software requirements engineering, starting from the original requirements gradually build complete specifications. In this paper, one of the methods of requirements specifications validation is given to confirm the accuracy of software requirements specifications and to fully describe the expected functions of software, so as to guarantee the requirements analysis in the field. The main contributions of this paper are listing as follows: (1) Propose a formal engineering method for requirement modeling of embedded control software. (2) Propose a special requirement description language, ATPRDL(Automatic Train Protection Requirement Description Language) for the field of train control, and models the corresponding requirements specifications. (3) According to the characteristics of embedded control software, we proposed a domain-specific dimensional analysis method of requirements oriented layer, and construct a domain-specific dimension system. (4) We developed a tool to support the practical application of the method, enabling engineers to conduct dimensional analysis quickly. By applying the method proposed in this paper, we experiment on the requirements of the real railway control software. The experimental results show that this method has a good usability. In the actual train control software development project, the defect detection ability and time efficiency are shown to be better than the traditional manual review methods.
AB - Embedded control software is the kernel component of today's control systems. For the safety-critical control systems, such as the railway system, aerospace control system and nuclear power control system, the function and safety of the control software embedded in them are directly related to the safety of people's lives and property. After long-term research and practice, both academia and industry realized that system requirements are the starting point of the software engineering life cycle and the fundamental way to ensure software quality. Therefore, how to ensure the correctness of software development process from the source has been an important research topic in the field of software engineering. However, in the software development of industry, the correctness of system requirements can only be guaranteed by manual review. Manual review of system requirements described in natural language is not only inefficient, but also unable to ensure the correctness of requirements. There is still a lack of a formal modeling analysis method to validate requirements. Requirements validation is an important engineering activity to ensure the quality of the formalized specifications constructed. As an important method of requirements validation, dimensional analysis can check out potential defects caused by wrong definition of dimension and reduce errors caused by improper use of dimensions. At present, the dimensional analysis in computer systems is mostly based on the code level and not applied in the requirements field, while some hidden errors in safety-critical embedded control systems are mostly caused by the wrong use of dimensions in requirements, which leads to software development errors from the source. Therefore, we propose a requirements modeling and dimension analysis method by according to the characteristics of the railway, which is based on formal engineering method. First, the requirements engineer converts the original requirements specifications into the requirements specifications described using the domain-specific requirements description language, then automatically extracts the requirements model using the tool. Finally, carries out the subsequent dimensional analysis on the basis of the requirements model. This work combines with the basic principle of software requirements engineering, starting from the original requirements gradually build complete specifications. In this paper, one of the methods of requirements specifications validation is given to confirm the accuracy of software requirements specifications and to fully describe the expected functions of software, so as to guarantee the requirements analysis in the field. The main contributions of this paper are listing as follows: (1) Propose a formal engineering method for requirement modeling of embedded control software. (2) Propose a special requirement description language, ATPRDL(Automatic Train Protection Requirement Description Language) for the field of train control, and models the corresponding requirements specifications. (3) According to the characteristics of embedded control software, we proposed a domain-specific dimensional analysis method of requirements oriented layer, and construct a domain-specific dimension system. (4) We developed a tool to support the practical application of the method, enabling engineers to conduct dimensional analysis quickly. By applying the method proposed in this paper, we experiment on the requirements of the real railway control software. The experimental results show that this method has a good usability. In the actual train control software development project, the defect detection ability and time efficiency are shown to be better than the traditional manual review methods.
KW - Dimensional analysis
KW - Embedded software
KW - Formal method
KW - Requirements modeling
KW - Requirements validation
KW - Software engineering
UR - https://www.scopus.com/pages/publications/85096357271
U2 - 10.11897/SP.J.1016.2020.02152
DO - 10.11897/SP.J.1016.2020.02152
M3 - 文章
AN - SCOPUS:85096357271
SN - 0254-4164
VL - 43
SP - 2152
EP - 2165
JO - Jisuanji Xuebao/Chinese Journal of Computers
JF - Jisuanji Xuebao/Chinese Journal of Computers
IS - 11
ER -