Methodology for Reusing Real-time HiL Simulation Models in the Commissioning and Operation Phase of Industrial Production Plants

Abstract

Automation Industry is dealing with productivity, costs, quality and flexibility in the manufacturing process. In addition to the classic requirements of increasing productivity, reduction of costs and increasing quality, flexibility plays a significant role in the modern manufacturing process. These requirements are fulfilled by development, adaptation, and application of advanced engineering processes and methodologies improving mechanics, electronics and software development. One methodology that is widely applied in the development of complex mechatronic systems is HiL Simulation. By definition, a Hil (HiL) Simulation is a system that combines parts of a real system with parts of its simulated counterpart (Stoeppler et al., 2005; Grega, 1999). It is one of the best-known methods for testing embedded systems (Yan et al., 2005; Boot et. al., 1999) by providing the possibility of validating software execution on the original hardware without the dangers that are evoked by a malfunction of the controller in combination with the real mechanic system (Pfister et al., 1999). In the early stages of the production plant lifecycle, ranging from Basic Design to Plant Engineering, discrete event simulation models are created reproducing the behavior of the mechanic system. Supporting the engineering of the control system, these models are executed in a real-time environment and connected to the control logic for desk checks. At the end of the Engineering Phase, the application of HiL Simulation shifts to the virtual commissioning of the overall process control system (Deutel & Suhm, 2002; Rohrlack, 2008). The commissioning of modern mechatronic production plants is a critical cost intensive phase. At the beginning of the commissioning phase, the controller and its embedded control software and its correct interaction with the production plant are validated by HiL Simulation. These techniques strongly base on accurate simulation models. Because of model inaccuracies between model and specification as well as between model and real plant, it is unlikely to find all errors in the controller software solely by controller testing with HiL Simulation. But even one failure in the controller may cause critical