Fault detection for automotive semi-active dampers

摘要

Advanced electronic monitoring systems are becoming widely used to assess faults of several automotive systems. In particular, the detection of suspension faults can be exploited to prevent vehicle ride handling problems. Advanced active and semi-active damping systems are also becoming widely used. In this context, on-line diagnostic systems are likewise indicated in order to detect fault states. However, fault detection schemes for these systems pose additional challenges because of the non-linear dynamics and the dependency on external control signals. One of the first obstacles for appropriate damper fault detection is the modeling of the fault, which has been shown to be of multiplicative nature. However, several detection schemes have been developed using additive faults because they are easier to handle mathematically. A fault detection scheme for magneto-rheological dampers is derived. The proposed scheme is shown to be able to detect multiplicative damper faults under a wide range of road-profile conditions using only acceleration sensors and low computational overhead. Several general and important characteristics applicable to all damper fault detection schemes are elucidated. Early results are promising.