Multivariable State-Space Identification of Dissolve Oxygen Control

Abstract

This work is proposed to identify a linear timeinvariant dynamic model of dissolve oxygen (DO) of a wastewater treatment plants with multilevel pseudo random signals as an excitation input. DO is always known as a main variable in wastewater control. For this purpose, a state-space model that emphasize on numerical subspace state-space system identification (N4SID) is applied. The works include the development of perturbation input signals, Identifying the estimation model continued by validating the model performances by Variance Accounted For and mean relative squared error. It was observed that the estimated model with multilevel input offers good predicted behavior’s compared to two-level pseudo random binary input signal. Benchmark Simulation Model BSM1 was applied as data generator for identification procedures.