On the design of adaptive derivative estimator using variable structure techniques

This paper is concerned with a novel kind of signal derivative estimators. Based on the concept of variable structure systems (VSS) and by introduction of adaptive techniques, the authors propose a new derivative estimator which possesses properties of filtering and obtaining derivatives of the observed signals in the presence of the measurement noise. The convergence of the proposed derivative estimator is guaranteed by means of the equivalent control. An adaptive algorithm is also incorporated to remove the requirement for the a priori knowledge about the boundary of the observed signal. The proposed derivative estimator shows the property of adapting to the noise parametric changes in an unpredictable way.