Fuzzy identification of systems and its applications to modeling and control

Analysis

Why This Paper Matters

This paper is a seminal contribution to fuzzy logic and control systems, introducing the Takagi-Sugeno (TS) fuzzy model. At a time when fuzzy logic was primarily used for rule-based expert systems, Takagi and Sugeno proposed a systematic method to identify fuzzy models from data, bridging fuzzy reasoning with linear system theory. The paper's significance is underscored by its over 20,000 citations, reflecting its foundational role in fuzzy modeling and control.

The TS model's key innovation is representing each fuzzy rule's consequence as a linear function of inputs, rather than a fuzzy set. This allows the overall model to be a weighted sum of linear subsystems, enabling the use of linear control design techniques for nonlinear systems. The paper also demonstrates practical applications, showing the method's effectiveness in real industrial processes.

Technical Contributions

• Fuzzy Model Structure: Each rule has a premise defining a fuzzy subspace of inputs and a consequence as a linear input-output relation.
• Identification Method: A systematic procedure to determine fuzzy partition of input space and linear parameters from input-output data.
• Applications: Demonstrated on a water cleaning process and a steel-making converter, showing practical feasibility.

Results

The paper presents qualitative results from two industrial applications: a water cleaning process and a converter in steel-making. No quantitative metrics (e.g., error rates, comparison benchmarks) are provided, but the applications illustrate the method's ability to model nonlinear processes effectively.

Significance

The Takagi-Sugeno fuzzy model became a cornerstone of fuzzy control and modeling, enabling the integration of fuzzy logic with classical control theory. It has been widely applied in robotics, automotive systems, and process control, and inspired numerous extensions and variants. This paper is a classic reference in the field of computational intelligence.