SELECTION OF AN OPTIMAL FORCE–STATE MAP

The restoring force method and the equivalent force–state mapping technique have been used to characterise non-linear mechanical systems. Both acquire a number of samples and process them to produce a non-parametric representation of the non-linear force as a function of two state variables. Errors are primarily introduced by an incomplete model and by measurement noise. By establishing a trade-off between both error sources it is possible to attain an optimal model. In this paper it is shown how such an optimal model is obtained by selecting the number of grid elements and their respective distribution. The optimal grid selection method is illustrated for automotive shock absorbers.