A Method of Dynamic System Identification over Computer Networks
Takehito Azuma · Review of applied physics · 2015
This paper discusses a problem about dynamic system identification over computer networks. A method to identify mathematical models of controlled objects over computer networks is proposed based on modified 1bit delta-sigma transformations and least squares methods for ARX models. The efficacy of the proposed method is verified based on numerical simulations and experiments. In the experiments, mathematical models of a DC motor are identified by using input-output signals of the DC motor over the developed computer network.