Development of a model predictive controller for an active torsional vibration damper to suppress torsional vibrations in vehicle powertrains

Abstract

The pressure of exhaust emission regulations on automotive manufacturers to reduce environmental pollution and fuel consumption of internal combustion engines (ICEs) have stimulated the works on the downsizing, downspeeding, and turbo supercharging concepts which result in boosted engine torsional vibrations. Despite significant momentum in the implementation of those concepts in modern ICEs in recent decades, similar progress has not taken place in parallel at torsional vibration isolation systems. To this end, this article centers on the development and implementation of a model predictive controller (MPC) on a novel active torsional vibration damper (ATVD) in which inertia, stiffness rate, and damping rate parameters can be varied to minimize torsional vibration transmission to the vehicle powertrain. Dynamic response of the ATVD is examined using an MPC inside a closed-loop control architecture with predicted variables. The MPC structure, state-space plant model, and physical constraint definitions are composed to be utilized in prediction models at various engine operating points. The MPC performance is evaluated in a co-simulation environment using Simcenter Amesim, NX Motion, and Matlab Simulink software, and are compared with that of the fuzzy logic controller (FLC). The simulation results clearly indicate that the MPC applied to the ATVD system has certain advantages over the FLC and is able to provide satisfactory isolation of the powertrain from engine-borne torsional vibrations while satisfying the physical constraints.

Research on noise control of automobile exhaust system under idling condition based on modal analysis

“…Without considering the influence of the torsional stiffness of the mount, the powertrain mount is simplified as a spring damping element in three directions, as shown in Figure 2. 13…”

Novel Powertrain Vibration Controller With Six Rules-Based Fuzzy Inference for Time-Fluctuated Control Period

“…Many control strategies have been investigated, including sliding mode control [7], [8], switching control of linear-quadratic regulator (LQR) [9], and model reference adaptive Linear Quadratic Tracking (ALQT) control [10]. Additionally, model predictive control (MPC) is widely used in many attempts [11]- [14]. The complicated properties of powertrain systems, such as nonlinearities and mechanical constraints, are however accommodated by standard MPC through computationally intensive optimization calculations and elaborate modeling.…”