Вісник № 01. Системний аналіз, управління та інформаційні технології
Постійне посилання колекціїhttps://repository.kpi.kharkov.ua/handle/KhPI-Press/47393
Переглянути
Документ Designing information support for evaluating the quality of embedded software(Національний технічний університет "Харківський політехнічний інститут", 2020) Zybin, Vladyslav Ivanovich; Liutenko, Iryna VictorivnaThis article presents a system for evaluating the quality of embedded software using a decision system based on fuzzy logic. These approaches will improve the assessment of software quality, due to its features. This article defines the main criteria for software quality used in assessing the quality of the software. The main literature was examined, in which fuzzy logic was described, decision-making systems using fuzzy logic, as well as software quality assessment systems, including software for embedded systems. The main characteristics and properties of embedded syst ems were considered. Based on the considered characteristics and properties of embedded systems, the ranking of criteria was made, which will be further used in the software quality assessment methodology. The main criteria that are used to evaluate the quality of software were considered, and the criteria presented were distributed according to the degree of influence on the assessment of the quality of software of embedded systems. Fuzzy logic was considered, and more precisely: the basic properties of fuzzy logic and fuzzy numbers, the basic mathematical operators applied to fuzzy numbers. The system for constructing rules for the rule base, as well as the defuzzification process, built on the basis of the centroid method, is analyzed. An example of software evaluation for embedded systems was considered. In this example, linguistic variables were determined, as well as their numerical ranges, which were used for the initial assessment of the quality criteria of this software. Each range of ratings was distributed according to the influence of a criterion on software quality. The output linguistic variable and its numerical value were also determined. In the end, based on the set values, an estimate of the set software was derived. The theoretical result obtained in this article is the basis for constructing a system for evaluating software quality for embedded systems.