Intelligent dynamic simulation of a lime kiln with linguistic equations

A new dynamic simulation tool for intelligent control design has been developed on the basis of Linguistic Equation (LE) approach and fuzzy logic. The fuzzy LE approach combines various information sources including mechanistic models, data-driven learning algorithms and expert knowledge to dynamic approximators, state space models and semi-mechanistic models. Understanding of the underlying phenomena and structure of the model are dealed with interactions, and nonlinearities of the system are described by membership definitions tuned by input-output data. Interactions, membership definitions and effective time delays depend on working point conditions. Delays are handled as smoothly changing fuzzy numbers. In the multimodel approach, the smooth switching between various submodels provides a robust platform for testing controller alternatives. The robust dynamic simulator is continuously used in developing adaptive industrial controllers.