Probabilistic optimization of a quarter car suspension with multiobjective framework and gradient based approximation
Palavras-chave:
multi-objective optimization, vehicle suspension, quarter car model, PSO, robust optimizationResumo
In this paper, a multi-objective robust optimization methodology is applied to the suspension
optimization problem of a quarter-car numerical model. In order to increase the driver’s comfort without
compromising the drivability, the chosen objective function was the weighted RMS acceleration according to
ISO 2631 with constrain regarding the suspension working space. The robust optimization is based in a
probabilistic approach, more advanced compared to the interval based approach. Monte Carlo simulations are
made to compare the statistics of the problem, as well as the failure probability. While the deterministic solution
found 3.97% better mean acceleration values when compared to the robust optimization, the chosen solution
generated by the multi-objective robust optimization results in a much lower failure probability: 10.55% for the
Robust against 50% for the deterministic.
