Uma análise empírica de abordagens para a geração de testes abstratos baseados em modelos de redes Petri coloridas

Abstract

The model-based testing (MBT) relies on models of the system’s behavior to generate test sequences. Testers usually reuse formal models designed using specification languages such as coloured Petri nets (CPN) and state machines to design tests for safety-critical systems. There is a considerable number of research studies focused on analyzing the usage of specification languages to conduct MBT. However, there is still a research gap regarding the analysis of test sequence generation approaches designed for CPN (e.g., state space- and simulation-based approaches). In this article, we present a comparative analysis of test sequence generation approaches based on CPN models by means of literature reviews and a case study on medical systems: electrocardiography and insulin infusion pump. The results presented in this article provide insights for testers who need to select the most adequate test sequence generation approach when applying the MBT using CPN. The state space analysis and exploration guided by model checking showed to be the most cost-effective approach during our empirical evaluation.