Map-based approach to fault detection for piecewise-affine systems

This paper addresses the problem of parameter estimation for fault detection using piecewise-affine models. Given a nominal model consisting of a partitioning into affine sections and assigned parameters for each section, the problem of fault detection is related to the problem of parameter estimation for the affine sections. In general, when a technical process is operated, only measurement data covering some sections are available. Furthermore, these data may not be suitable for estimating all parameters well if they do not cover the whole domain of the corresponding affine section. In this paper, a new approach is presented which uses a data buffer to overcome these limitations.