Automatic Transformation of UML System Models for Model-based Error Propagation Analysis of Mechatronic Systems

Abstract

Abstract: Mechatronic systems consist of heterogeneous components: mechanical parts, hardware, and software. Appropriate models, which describe the mutual physical interaction on common Abstract levels, are required. UML is a widely accepted candidate for design and model-based analysis of the mechatronic systems. For the error propagation analysis on system level we have introduced a stochastic dual-graph error propagation model. This model captures control and data flow aspects of the system and allows the computation of various reliability metrics using discrete time Markov chain models. In our recent case-studies, UML Activity Diagrams have been used as baseline models. However, the transformation process was not fully automatic. This process is not so straightforward, despite the obvious structural similarities of the activity diagrams and our error propagation models. This article presents a new fully automatic method for the transformation of annotated activity diagrams. The transformation algorithm is described in detail with formal set-based mathematical notations. The article addresses both theoretical and technical sides of the problem. The method is demonstrated as a part of a complete analytical workflow in the frame of a mechatronic case study.