Optimal Recursive Identification Techniques from Quantized Outputs

GÓMEZ, JUAN CARLOS; SAD, GONZALO DANIEL

Buenos Aires

Congreso; 28 Congreso Argentino de Control Automático (AADECA2023); 2023

Recursive identification algorithms for the parameter estimation of linear systems from multilevel quantized outputs are introduced in this paper. The proposed algorithms are proved to be optimal in the sense that they minimize the a posteriori parameter estimation error covariance matrix. Numerical simulations are carried out to illustrate their performance for different quantization steps and Signal-to-Noise Ratios, as well as their capability to track time varying parameters.