Identiication of Mimo Systems by Inputtoutput Ts Fuzzy Models
نویسنده
چکیده
|A number of techniques have been introduced to construct fuzzy models from measured data. Most attention has been focused on multiple-input, single-output (MISO) systems. This article concentrates on the identi cation of multiple-input, multiple-output (MIMO) systems by means of product-space fuzzy clustering with adaptive distance measure (the Gustafson-Kessel algorithm). The MIMOmodel is represented as a set of coupled input{output MISO models of the Takagi-Sugeno type. Knowledge of the physical structure can easily be incorporated in the structure of the model. Software implementation in the form of a Matlab toolbox is brie y described. A simulation example of four cascaded tanks is given. Keywords|Fuzzy modeling, nonlinear identi cation, multivariable (MIMO) systems, fuzzy clustering, Matlab.
منابع مشابه
SECURING INTERPRETABILITY OF FUZZY MODELS FOR MODELING NONLINEAR MIMO SYSTEMS USING A HYBRID OF EVOLUTIONARY ALGORITHMS
In this study, a Multi-Objective Genetic Algorithm (MOGA) is utilized to extract interpretable and compact fuzzy rule bases for modeling nonlinear Multi-input Multi-output (MIMO) systems. In the process of non- linear system identi cation, structure selection, parameter estimation, model performance and model validation are important objectives. Furthermore, se- curing low-level and high-level ...
متن کاملFuzzy Control of Multivariable Processes
Most processes in industry are complex mul-tivariable systems with nonlinear and time-varying dynamics. These processes are often diicult to model. However, often much expert knowledge and process data is available. Fuzzy modeling and control techniques can eeciently integrate this information. However, most of the systems considered in literature are of low-order and SISO or MISO systems. For ...
متن کاملStability Analysis of Multivariable Fuzzy Control Systems Using the FAST Toolbox
This paper studies the stability analysis of multivariable fuzzy control systems. Particularly, the stability of Takagi-Sugeno (TS) systems is addressed. The results presented in the paper make possible the application of both input-output and frequency response methods to TS systems identified from Input-Output data. The paper also presents the Fuzzy Algorithm Stability Tool (FAST) toolbox dev...
متن کاملFuzzy Predictive Control of Fractional-order Nonlinear Discrete-time Systems
At the end of the 19th century Liouville and Riemann introduced the notion of a fractional-order derivative, and in the latter half of the 20th century the concept of the so-called Grünewald-Letnikov fractional-order discrete difference has been put forward. In the paper a predictive controller for MIMO fractional-order discrete-time systems is proposed, and then the concept is extended to nonl...
متن کاملOn Observer Design Methods for a Class of Takagi-sugeno Fuzzy Systems
The generalized design principle of TS fuzzy observers for one class of continuous-time nonlinear MIMO systems is presented in this paper. The problem addressed can be indicated as a descriptor system approach to TS fuzzy observers design, implying the asymptotic convergence of the state observer error. A new structure of linear matrix inequalities is outlined to possess the observer asymptotic...
متن کامل