Genetic algorithms in fuzzy model inversion

Recently model-based techniques became very popular and widely used in solving measurement and control problems. For measurement data evaluation and for controller design also the inverse models are of considerable interest. The inverse models can be utilized either as a direct compensation of some measurement nonlinearity, or as a controller mechanism for nonlinear plants. In this paper an improved technique for fuzzy model inversion is introduced. Multiple-input single-output (MISO) forward fuzzy models are considered, where inversion is performed simply by interchanging the role of the output and one of the inputs. The proposed method is based on a simple nonlinear state observer, which reconstructs the selected input of a system, represented by a forward fuzzy model, from its output and the remaining inputs using an appropriate prediction-correction type control strategy and a copy of the fuzzy model itself. The overall performance of the suggested technique is highly influenced by the nature of the nonlinearity and the actual prediction-correction mechanism applied. The novelty of this paper is the introduction of genetic algorithms to control the iterative model inversion.