Comparative analysis of different semi-active suspension control techniques applied on real world MR dampers
Vehicular Dynamics, Vibration control, Semi-active damper, Magnetorheological damper, Genetic optimization.
This thesis aims to use simulations for evaluating the performance of different semi-active suspensions control strategies at maximizing the comfort and safety of vehicle systems equipped with magnetorheological (MR) dampers with realistic behaviors.
With the objective of evaluating the performance of diverse controllers in a realistic vehicle plant, many MR damper models were considered. Four different proposals were evaluated in Simulink: The Bingham body model, the asymmetric Bouc-wen model, a high performance Double Sigmoid model referred to as the Wang model, and a novel modified form of the Wang model based on the logistic equation. Each model parameters were optimized by a genetic algorithm and ranked according to how well they fit Force data from experiments. The modified Wang model proposed in the thesis performed best. Modifications were done to the ”peak velocity” parameter in the Wang model due to divergence issues in vehicle simulations. Two new calculation methods were proposed. Simulations comparing the techniques revealed the proposed formulations were stable under a wider range of simulation conditions. One was more precise for harmonic disturbances, while the other was better for transient disturbances. The latter was chosen. The modified Wang model was used in the vehicle simulations.