Digital Control Systems Design, Identification and Implementation电力—5 System Identification The Bases.pdfVIP

5 System Identification: The Bases In this chap ter the basic pri ncipl es of identif ication of dynamic systems are first introduced. This is followed by a p resentation of the main types of parameter adaptation algorithms used in recursive identification methods. The choice of inpu t signals for identification and the influ ence of disturbances is also discussed. The last section presents the general structure of recursive identification methods. 5.1 System Model Identification Principles Identification means the determination of the model of a dynamic system from input/output measurements. The knowledge of the model is necessary for the design and the implementation of a high performance control system. Figure 5.1 sums up the general principles of controller design. In order to design and tune a controller correctly, one needs: 1. To specify the desired control loop performance and robustness 2. To know the dynamic model of the plant to be controlled (also known as the control model) which describes the relation between the control variations and the output variations 3. To possess of a suitable controller design method enabling to achieve the desired performance and robustness specifications for the corresponding plant model The notion of the mathematical model of a system or phenomenon is a fundamental concept. In general, a multitude of model types exist, each one dedicated to a particular application. For example, the knowledge type models (based on the laws of physics, chemistry, etc...) permit a fairly complete system description and are used for plant simulation and design. These models are in general extremely complex and