Fault Diagnosis of Driveline System Using Response Optimization

This paper describes an optimization approach for damage detection in automobile driveline systems. In operating sequence of an automobile, the vibration response at any point along the driveline should remain same, but practically it changes due to driveline faults. Usually damage is measured in terms of reduction in the stiffness of the structure. In this paper, initially simulated torsional vibration response data is obtained along the driveline by considering stiffness reduction coefficients in each element. A residual forcing function is formulated from the equations of motion and the unknown stiffness reduction coefficients corresponding to the present state of response are predicted by minimizing the defined residue. Effectiveness of genetic algorithms in solving such problems is illustrated. The methodology is presented with help of a driveline system of a light truck.