An Efficient High-Order Time-Step Algorithm With Proportional-Integral Control Strategy for Semirecursive Vehicle Dynamics

Abstract

Dynamics of complex mechanical systems can be modeled and solved efficiently, often even in faster-than-real-time by employing semi-recursive formulations and their various versions. To simulate complex mechanical systems accurately and efficiently, an optimal combination of a multibody formulation and a numerical time integration scheme is critical. To this end, this paper introduces a procedure based on the double-step semi-recursive formulation and adaptive time-step algorithm. The introduced adaptive time-step is based on proportional-integral controller that is particularly suitable for efficient computation of vehicle dynamics. The introduced proportional integral control strategy accounts the local truncation error and recent alteration concerning the local truncation error which improves the numerical stability and efficiency. As numerical examples, a 15-degree-of-freedom sedan vehicle model and a 17-degrees-offreedom light truck model are analyzed by using the double-step semi-recursive formulation with an adaptive time-step algorithm. The results highlight the efficiency gain of the presented adaptive time-step algorithm. It can be concluded that the introduced approach is suitable for efficient simulation of highly nonlinear dynamics model.

Validation of a Real-Time Capable Multibody Vehicle Dynamics Formulation for Automotive Testing Frameworks Based on Simulation

“…Figs. [14][15][16] show that the simulation results match the data obtained with the real test vehicle, providing very low RMSE values ( Table 7). In conclusion, the low RMSE obtained for all the tested maneuvers demonstrates the optimal accuracy of the pro-…”

“…These compensate the high computational cost derived from the use of game engines [19] to simulate complex environments and on-board sensors, with a simplified vehicle dynamics model. On the other hand, multibody formulation based models [11]- [14], can represent more accurately vehicle dynamics, but present an increased computational cost that limits their real-time implementation [21], [22], especially if flexible bodies are implemented [20]. Likewise, due to this high computational cost, computationally powerful devices would be needed in order to fulfill the real time requirements.…”