This paper presents a theoretical proposal of fuzzy instrumental variable (FIV) methods for nonlinear discrete time systems identification and parameter estimation in noisy environment based on the fuzzy instrumental variable concept. The theoretical analysis is presented using a suitable formulation associated to Takagi-Sugeno (TS) fuzzy model. The complexity of the algorithm is quite low and the statistical properties show that the assintotic error of the parameter estimates go to zero as the data series length increases.
System identification; Fuzzy systems; Fuzzy instrumental variable
