FMI to facilitate virtual development in industrial automationAdina Tunér
FMU based workflows demonstrated in control design for B&R Automation
Initiated in the automotive industry to facilitate virtual development processes, simulations based on FMI are required to get innovations to the market fast. After the automotive industry, other fields where computer-aided engineering plays an important role are now keen to accelerate product development. Model-based design workﬂows using FMI offer a promising future in this direction.
Recent industrial MSc thesis projects involving B&R Automation, Lund University and Modelon delivered contributions to defining model-based design workﬂows based on the Functional Mock-up Interface (FMI) involving Software-in-the-Loop (SiL) and Hardware-in-the Loop (HiL) simulation.
The workflow in these projects included plant modeling in Dymola (Modelica), control design in Matlab/Simulink, implementation of the control system in B&R Automation Studio, and running the full system using B&R PLC and physical process, as shown in Fig 1.
Fig.1 FMU-based workflow: the physical model of the pendulum carried through multiple tools
The main process used to test the workflow was the Inertia wheel pendulum, see Fig 1. The control objective is to swing the pendulum to an upright position and hold it there, using the inertia wheel. The projects had access to the physical process as well as a PLC connected to it, both from B&R Automation.
In a first thesis, a study on the Evaluation of FMI-based Workﬂow for Simulation and Testing of Industrial Automation Applications proposed a SiL workflow for simulating the B&R Reaction Wheel Pendulum (see Fig.2), with work conducted in Automation Studio using a model imported with FMI.
Fig.2 Inertia wheel pendulum controlled by a PLC – an excellent system from B&R Automation for testing a new workflow based on the newly developed FMI Support in Automation Studio.
Fig.3 Virtual model of the pendulum developed in Dymola using components from the Modelica Standard Library.
In a second study, Model-based design of industrial automation solutions using FMI, a HiL simulation of the pendulum was performed, thus showcasing the strength and possibilities of using FMI in testing. Comparing the SiL and HiL systems showed only small differences in terms of pendulum swing-up and settling times, showing the strength of the workflow.
From these first contributions we can see that FMI-based workflows prove to offer a promising basis for virtual development in industrial automation.
To learn more on how to develop virtual models for your product or system, please contact firstname.lastname@example.org.
- Gunnarsson, S., Evaluation of FMI-based Workﬂow for Simulation and Testing of Industrial Automation Applications, MSc Thesis, Lund University, 2016
- Erwall, C., Mårtensson, O., Model-based design of industrial automation solutions using FMI, MSc Thesis, Lund University, 2016