Mikelsons, Lars; Ji, Hongchao; Brandt, Thorsten; Lenord, Oliver:
Symbolic Model Reduction Applied to Realtime Simulation of a Construction Machine
In: Proceedings of the 7th International Modelica Conference, Como, Italy, 20-22 September 2009 - Linköping: Linköping University Electronic Press, Linköpings universitet, 2009 - (Linköping Electronic Conference Proceedings), S. 765 - 774
Buchaufsatz/Kapitel in Sammelwerk2009Maschinenbau
Fakultät für Ingenieurwissenschaften » Maschinenbau und Verfahrenstechnik
Symbolic Model Reduction Applied to Realtime Simulation of a Construction Machine
Mikelsons, LarsLSF; Ji, Hongchao; Brandt, ThorstenLSF; Lenord, Oliver


The vehicle response of construction machines strongly depends on the tuning of the control system in interaction with the drive system. A compromise between performance and comfort needs to be found to fulfill the operators requirements on a high usability of the machine. In order to achieve an optimal behavior Hardware-in-the-Loop simulation techniques offer a suitable approach to determine the overall behavior in advance. Prerequisition is a realtime capable simulation model of the considered system. Therefore, in this paper the mathematical model of the system is automatically adapted by symbolic model reduction algorithms in order to match real-time requirements on a given hardware. Inputs to the automatic reduction algorithm are the complex mathematical system model, the desired realtime cycle and the number of floating operations per second (flops), which can be realized by the chosen target hardware. The outputs of the algorithm are the automatically reduced model, which is guaranteed to run in realtime on the target hardware and the maximal model error for the test scenario. In this paper, the reduction procedure is demonstrated for the complex hydromechanical model of a so-called kid steer loader. Summarizing, the proposed procedure of symbolic model reduction helps to reduce the developing phase of mechatronic prototypes dramatically as the adaptation of the system model with respect to the target hardware is completely automated.