Fault Estimation Observers for the Vehicle Suspension with a Varying Chassis Mass
Abstract
This paper proposes a descriptor Nonlinear Parameter Varying (NLPV) observer
for damper fault estimation in vehicle semi-active suspension systems, here considering the
highly practical case of a varying sprung mass, which is an open problem. This mass and the
control input are treated as known scheduling parameters for the design and implementation
of the observers, thus making the structure adaptive. This novel modeling leads to parameter-dependent dynamics and output matrices in the nonlinear parameter-varying formulation. To solve this problem, we extend an existing reduced-order observer design method (following descriptor system modeling) to this new case. Both polytopic and grid-based methods are then considered to compute and implement the problem solutions. The designs are then assessed using frequency-domain analysis and realistic time-domain simulations.
Origin : Files produced by the author(s)