Identification of first-order plus dead time using typical test signals

In this paper, we propose an online continuous identification technique of first-order plus dead time (FOPDT) using typical inputs. The impulse test is assessed at the beginning to deduce the basis of the identification algorithm, due to its uncomplicated physical interpretation. The remaining typical test signals are then inferred easily. These test procedures lead to a mathematical formulation allowing the simultaneous online identification of time delay and the parameters of FOPDT model. The robustness property of the proposed method is enhanced due to the convolution by the sigmoid function. The identification problem is solved using the Levenberg-Marquardt algorithm. Numerical simulations are given at last to illustrate the validity of the theoretical findings.