The time-frequency localization and multiresolution properties of wavelets are descussed in context of control applications. This paper provides an overview of wavelets from a process control perspective. Various control problems where wavelets could be particularly advantageous are discussed in this paper. The purpose of this paper is to discuss the wavelet transforms and their possible applications in process control which have already been studied or being investigated currently. Process control problems often involve dynamics on time-scales. Different outputs may also exhibit behaviors on different time-scales. In these cases the sampling rate for each input and output would be chosen to be commen rate with the associated time-scales. In this paper addresses the model predictive control, presents a wavelet transformation of the model predictive control problem. The purpose of the presentation is to illustrate how the time-frequency interpretations of the inputs and outputs can be of advantage in control design. The wavelet theory from the process control perspective is presented here. A brief overview of the existing results and possibilities for future applications are gaven in the paper. The research in this area is still in rudimantary stages. Most of the applications in process engineering are yet to be explored and the realization of the full power of the wavelet transforms will take some more time.
展开▼
