Grey box-based inverse wagon model to predict wheel–rail contact forces from measured wagon body responses

Abstract

A grey box-based inverse wagon model was developed to estimate wheel–rail contact forces using only measurements of wagon body responses as inputs. The project is based on a similar application using a deterministic inverse wagon model [F. Xia, C. Cole, and P. Wolfs, An inverse railway wagon model and its applications, Veh. Syst. Dyn. 45 (6) (2007), pp. 583–605]. The deterministic (i.e. white box) inverse wagon model has some limitations, one of which is that the high-frequency components of the forces at the wheel–rail interface cannot be predicted from the lower frequencies at the wagon body. They are effectively isolated by the suspension. To overcome this problem, the present research focuses on developing of a grey box-based inverse wagon. The high-frequency forces may be considered as the responses to random track irregularities with high frequency and appear as a random feature. The grey box-based inverse wagon is used to predict wheel–rail contact forces using only lateral and vertical accelerations, roll, pitch and yaw angular velocities measured on the wagon body.