Set-Bounded Joined Parameter and State Estimation for Model Predictive Control of Integrated Wastewater Treatment Plant Systems at Medium Time Scale

Abstract

The paper describes set-bounded parameters and state estimation component for Model Predictive Control of integrated wastewater treatment plant at medium time scale purposes. This is one of the components within Intelligent Hierarchical Control Structure designed in order to cope with problems of the controlled system such us: multiple time scale, highly non linear dynamics, small number of on-line measurements etc. The set-bounded joined parameter and state estimation algorithm generates robust estimates of these parameters and state for Model Predictive Control purposes. Both, the estimation algorithm and the Model Predictive Controller utilise a grey-box model of controlled plant dedicated for medium time scale. The set-bounded joined state and grey-box model parameters estimation algorithm is validated by simulation. The data were taken partly from real data records gathered at Kartuzy WWTP in Poland and partly from the Kartuzy plant simulation based on recently developed and calibrated plant model.