An integrated identification methodology for control and identification

An integrated identification methodology (IIM) is presented that appropriately combines expert systems and adaptive identification technologies for process control systems. Since many physical systems have both numerical and non-numerical characteristics, both qualitative and quantitative descriptions of the characteristics are proposed in the IIM for identification and control. Characteristic descriptions and parameters of the controlled process are identified using expert systems and adaptive algorithms, respectively, in order to meet a full-scale identification requirement. The proposed methodology is currently implemented in a Lisp based expert systems shell for synthesis, analysis and design of process control systems. In addition, examples, on-going work and insights concerning this issue are addressed.