RFDF design for a class of time delay systems with unknown inputs and parameter uncertainties using bicausal transformation

This paper tackles with the design of robust fault detection filter (RFDF) for a class of linear time invarient (LTI) systems with constant output state time delays including the effects of both parameter uncerttainties and parameter uncertainties. The approach used is to remove the time delays of the system i.e. to transform it to a delay free system by using bicausal change of coordinates approach. Then, design a RFDF using a multiobjective optimization technique for delay-free linear time invarient system which is equivalent to the original time delay system. The design problem of RFDF is formulated as a standard H_∞ model matching problem and is solved by using H_∞, optimization LMI techniques. For the detection of the fault, an adaptive threshold which depends on the inputs is determined finally. In the end, a numerical example is used to demonstrate the validity of the design approach.