Data Dependency Analysis in Industrial Systems

Sammanfattning: Software in modern industrial systems may have complex data dependencies. As a result of this, it can be hard for system developers to understand the system’s behavior, if they cannot explicitly see these dependencies. This thesis addresses this problem, with an emphasis on dependencies among data in systems built using the IEC 61499 standard. An analysis method was developed, with which we are able to extract data dependency information from basic and composite function blocks. The first step of handling this problem is to investigate how the data dependencies occur in IEC 61499. The second step is to create a formal definition of IEC 61499 elements that were needed in order to formulate the analysis method. Next, we define a dependency matrix, in which we store the information regarding dependencies between input and output data ports. Later, we formulate the necessary algorithms for data dependency analysis in basic and composite function blocks. Finally, the last piece of the puzzle is to develop a plug-in for Framework for Distributed Industrial Automation and Control – Integrated Development Environment. This plug-in is used to show that the analysis method is efficient and that the proposed analysis is applicable to IEC 61499 systems.