Vehicle suspension force and road profile prediction on undulating roads

Abstract

Controllable suspension systems have the capability of changing suspension forces. One control approach is to define a cost function with the aim of optimising either ride comfort or handling. Such controllers are usually reactive and not pro-active. Controllers can benefit significantly by having a priori knowledge of the effect that changing the suspension settings will have on the suspension forces. This is especially true for vehicles traversing very rough terrain. This paper addresses the a priori knowledge needed by predicting what the suspension forces will be before changing the suspension setting. The proposed approach involves estimating sprung and unsprung mass acceleration, estimating the road excitation, and then predicting the suspension forces. A quarter car model is used to illustrate the concept. Thereafter, the concept is extended to a nonlinear multibody dynamics model and finally validated experimentally. Results indicate that the suspension force in a different suspension modes can be predicted before switching suspension modes.