State Based Control of Wastewater Treatment Plants--Evaluation of the Algorithm in a Simulation Study

Abstract

This paper presents the design of a supervised controller with the primary objectives of improvement of process stability and reduction of operating costs. The controller uses a rule set for determination of the process state extracted from the results of a fuzzy c-means clustering algorithm. Additionally a finite state machine is used for the evaluation of identified states in order to check logical consistency and process behavior. The result of the state machine is a symbolic value which describes the condition of the process. Based on the condition-identification an appropriate control strategy is used. The supervised controller methodology is evaluated in a simulation study. A significant improvement of process stability and energy saving has been achieved.