Application of Fuzzy Transform for noise reduction in helicopter model identification

Abstract

This paper deals with the application of Fuzzy Transform in Control System Identification. The fuzzy transform method is design as a pair of two transform. The first one maps a continuous function to a real vector. The second one, inverse transform, is naturally a mapping from the space of real vectors to the space of continuous function. Later on, more and more applications had been appearing to be appropriate because of several valuable properties which are typical for the method. This paper proposes a short overview of these valuable properties. It however, avoids technical and mathematical details since they are not crucial for such a technical paper. Our main task is to introduce the method to practitioners, present its applications for smoothing of measured courses of output variables of control systems. Processed courses are used to control system identification. An example of Humusoft helicopter model identification is described in the paper with the purpose to show how to apply the Fuzzy Transform for smoothing of functions.