PAR: A Practicable Formal Method and Its Supporting Platform

• Published in: Formal Methods and Software Engineering Vol. 11232; pp. 70 - 86
• Main Authors: Xue, Jinyun, Zheng, Yujun, Hu, Qimin, You, Zhen, Xie, Wuping, Cheng, Zhuo
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2018; Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Formal methods; Formal specification; Model transformation; Modeling language; Par platform
• ISBN: 3030024490; 9783030024499
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-030-02450-5_5

The use of formal methods can significantly improve the reliability, correctness and efficiency of software development. Although formal methods has been invented for more than 40 years, but academia and industry do not have a unified understanding of what are formal methods and its essential characteristics. Formal methods has not been recognized and widely applied by academia and industry. The authors of this paper have long been engaged in the study of the essential features of Formal methods. The authors propose a new definition: Formal methods are a strict technology based on mathematics and tool support for software and hardware system, including high-level abstract specification, modeling language and different levels of model transformation tools. Based on this definition, this paper develops a practicable formal methods and its supporting platform, called PAR method and PAR platform, short for PAR. PAR consists of the following elements: requirement modeling language SNL, algorithm modeling language Radl, abstract program modeling language Apla, a set of rules for the model transformation and a set of automatic transformation tools from requirement models to algorithm models, to abstract program models and to executable programs. The goal of the transformations is to generate executable program. The elements embody 6 innovative ideas given in Sect. 2. There are two kinds of applications of PAR. One is that many nontrivial algorithms and programs have been developed formally. Another is formal developing several safety-critical information systems.