Probability-based global state detection of complex technical systems and application to mobile working machines

Frequently, the increasing level of automation requires a systematic consideration of numerous interacting components influenced by internal feedback mechanisms as well as interactions with human operators under varying environmental conditions. This places demands on modeling, which in general cannot be satisfied by traditional modeling concepts. In this paper, a model approach for complex technical systems is developed, which consists of two language layers. The first layer takes the physical knowledge of the technical system into account. The second one uses stochastic models in combination with superordinate higher-level terms to formulate properties concerning the overall system. By means of a case study dealing with the automation of mobile working machines, it is shown that this approach can be used to recognize driving situations under varying environmental conditions.