Mid Sweden University

Today the reliability of electronic systems strictly depends on the correct operations of the sensor involved in the normal use. This fact is more evident in applications where the security and safety of end-users are involved, reason why a suitable Instrument Fault Detection Scheme (IFD) including also Isolation feature (IFDI) becomes fundamental. The automotive field is one of the main areas where an IFDI scheme is mandatory. As an example, systems designed to on-line adapt and to electronically control the suspensions of motorcycles are today of great interest for motorcycle and after-market manufacturers. By the way meanwhile semi-active suspension systems can drastically improve the comfort and the traction performance of motorcycle, a promptness detection of faults involving this kind of system is fundamental for the safety and performance of the motorcycle. With this aim, the paper proposes an Instrument Fault Detection and Isolation (IFDI) scheme using analytical redundancy for the fault diagnosis of the front and rear stroke suspension sensors. Both suitable mathematical links and soft sensors based on artificial neural networks are proposed for the residuals generation in order to design and validate the proposed IFDI scheme. © 2020, Springer Nature Switzerland AG.