Networked Control with Time Delay Compensation Scheme Based on a Smith Predictor for the Activated Sludge Process

Abstract

In thispaper, a Smith predictor compensation scheme is developed to mitigate randomdelays in a networked wastewater control. The focus here is the process ofDissolved Oxygen (DO) concentration as part of the Activated Sludge Process(ASP). The networked wastewater control is a new method of wastewater controlin which the controller and the wastewater treatment plant are separated by a widegeographical distance and so a communication medium is required between them.The effects of network induced time delays on the DO process are investigated. Thesimulation results reveal that communication drawbacks adversely affect thestability of the closed loop DO process resulting in water odour, flocformation, poor sludge formation and eventual low effluent quality. This isbecause low DO concentration due to communication drawbacks leads to depletionof oxygen available to microorganisms that are meant to clean up thewastewater. A Smith predictor is therefore proposed to compensate for thesecommunication drawbacks. The nonlinear DO model is linearised using theinput/output feedback linearization method. PI controller is designed for thelinearised DO process while a Smith predictor scheme is proposed to eliminate timedelays in the control system.  Analyticaland simulation results confirm the effectiveness of the scheme over the PIcontroller to provide robustness for the control of the input-output feedbacklinearized DO process under the influence of random communication delays.