DE GRUYTER OLDENBOURG at – Automatisierungstechnik 2017; 65(12): 851–866 Methoden Santiago Soler Perez Olaya*, Stefan Mätzler, and Martin Wollschlaeger Reliable software-based control as enabler for flexible production systems Zuverlässige softwarebasierte Regelung als Baustein für flexible Produktionssysteme https://doi.org/10.1515/auto-2017-0089 Received August 25, 2017; accepted November 8, 2017 Abstract: The current evolution of the industrial produc- tion systems to cyber physical production systems requires an increased fexibility of the system structure that is nowadays still difcult to fnd in the industrial systems. The control applications are extremely strict by requit- ing jitter-free communication of sensor and control values in networked control systems. The software-based control approach presented here enhances the reliability of the control system using a control value matrix as information source. This approach benefts of predictive control algo- rithms that rely on model-based strategies. Keywords: Software-based control, model-based control, NCS, reliability, cyber physical production systems. Zusammenfassung: Die gegenwärtige Entwicklung der in- dustriellen Produktionssysteme hin zu Cyber-physischen Produktionssystemen erfordert eine fexiblere System- struktur, die derzeit in industriellen Systemen immer noch schwer zu fnden ist. Regelungsanwendungen erhöhen die Anforderungen, indem sie eine jitterfreie Kommunikation von Sensor- und Steuerwerten in vernetzten Steuerungs- systemen erfordern. Der hier vorgestellte Ansatz zur soft- warebasierten Regelung erhöht die Zuverlässigkeit des Re- gelungssystems durch Verwendung einer Steuerwertma- trix als Informationsquelle. Dieser Ansatz proftiert von *Corresponding author: Santiago Soler Perez Olaya, Chair for Industrial Communications, Technische Universität Dresden, Faculty of Computer Science, Institute of Applied Computer Science, Nöthnitzer Str. 46, Dresden, Germany, e-mail: santiago.soler@tu-dresden.de Stefan Mätzler, Martin Wollschlaeger: Chair for Industrial Communications, Technische Universität Dresden, Faculty of Computer Science, Institute of Applied Computer Science, Nöthnitzer Str. 46, Dresden, Germany prädiktiven Regelunggsalgorithmen, die auf modellbasier- ten Strategien beruhen. Schlüsselwörter: Softwarebasierte Regelung, modelba- sierte Regelung, NCS, Zuverlässigkeit, Cyber-physische Produktionssysteme. 1 Introduction Modern automation and control systems (AT) face the inte- gration of information technology (IT) at various levels of production systems’ hierarchies. While an integration in the upper levels – ranging from Enterprise Resource Plan- ning down to Manufacturing Execution Systems (MES) is rather straightforward, specifc requirements have to be considered at the Control Level and the Field Level. Those levels are characterized by dedicated, industry-specifc systems and components, explicitly designed to match the requirements. These requirements include availability, ro- bustness and resilience, and specifcally the time-related behavior in terms of real-time, end-to-end latency bound- aries, or maximum jitter. The term Quality of Service (QoS) has been introduced to address such requirements. Besides addressing systems and components, this is also valid for the communication systems, and for soft- ware that is applied at these levels. It is expected to op- erate such systems in a 24/7 mode, sometimes running uninterrupted for several years. This rises requirements for engineering and maintenance of these systems as well. Traditional feldbus systems, the network and the feld devices, are engineered to optimally fulfll a certain application. With the emerging development of electronics and IT systems, driven by consumer market requirements, not only the functionality of electronic components has been grown signifcantly, but also an enormous cost reduc- tion has been achieved. These were the main reasons to adopt IT technology also for automation and control tasks. Brought to you by | Göteborg University - University of Gothenburg Authenticated Download Date | 12/17/17 9:29 AM