Estimation of Suspension Parameters for the MOWAG Piranha IIIC 8WD Combat Vehicle using the Markov Chain Monte Carlo Method

Resumo

This study presents a methodology for estimating dynamic suspension parameters of the MOWAG Piranha IIIC 8×8 armored vehicle used by the Brazilian Navy. A simplified computational model representing 1/8 of the full system was developed with two degrees of freedom (2-DoF), corresponding to the suspended mass (chassis) and unsprung mass (wheel–tire), connected by a spring–damper system. Equations of motion derived from Newton’s second law were solved numerically in MATLAB® to simulate vertical responses over a standard trapezoidal obstacle. Experimental chassis vertical acceleration data were used in an inverse problem to estimate spring stiffness (ks) and damping coefficient (c) through Bayesian inference using the Markov Chain Monte Carlo (MCMC) method. An initial validation step employed the technique known as inverse crime, in which synthetic measurements generated by the same forward model were used to verify the algorithm’s ability to recover known parameters. A sensitivity analysis using finite difference approximations showed that both parameters significantly influence the system’s vertical response. Results demonstrate that ks and c play a critical role in suspension dynamics and that the inverse crime procedure ensured correct MCMC implementation before applying it to experimental data. This validates the proposed methodology and highlights the potential of Bayesian inference for parameter identification in complex dynamic systems.