A nonlinear model to predict drivers’ track paths along a curve

This study proposes a prediction model about the trajectories a vehicle,in isolated conditions,along a curve of a road. as we know,the road environment induces stress on users and,under certain conditions,influences driving behavior. it is of advantage then,to isolate and identify those conditions from among the numerous variables,which are actually the most significant so as to prevent or mitigate the occurrence of dangerous maneuvers. on the basis of an experiment performed using an instrumented vehicle,we collected a data base to which we subsequently applied neuro-fuzzy techniques for the selection of the most representative variables. we then used these data to prepare a nonlinear dynamic hammerstein-wiener’s model able to predict the track paths along curves. the findings were encouraging since almost all the results obtained from the validation checks proved satisfactory. this research is the first step in the identification of complex systems and could be applied in road safety measures and design of new and existing roads. © iran university of science and technology 2016.