In this dissertation a model-based test case generation approach to test the reaction of automated production systems (aPS) to technical process and system faults is presented. Proposing a lightweight notation and thereby enabling industrial companies to specify a test model for testing the reaction of aPS to faults is one main focus of the dissertation. The model chosen is the timing diagram as both technical process engineers as well as aPS software engineers comprehend and work with this notation in engineering practice. One highlight of the dissertation lies in the test generation process, which takes the adapted timing diagram as a basis and combines it with a classification of faults as the programmable logic controller (PLC) perceives them. In this way, test cases to test the reaction of aPS to faults are generated. The approach aims at checking whether an error handling routine for the fault is present and correctly implemented. Another contribution is the evaluation of the feasibility of the software-implemented fault injection to test aPS PLC software. In order to inject the faults, PLC programming environments and program execution have been analyzed and a concept for fault injection has been developed. The execution as Hardware-in-the-Loop is realized using commercially available tools. Modelling support for the approach and the test case generation has been realized within a prototypical editor. The test case execution has been realized within a commonly used PLC programming environment including the function of importing the test cases from the prototype. Consequently, model-based test case generation from timing diagrams can directly be realized.weiterlesen