MULTIVARIATE CONTROL PERFORMANCE ASSESSMENT OF MPC IN WASTEWATER TREATMENT PROCESS

摘要

As the public awareness of environmental protection increases and the environmental regulations become more stringent, effective control of the wastewater treatment process (WWTP) has become a research hotspot. However, with strong interaction between variables it is difficult to control the WWTP since the process is nonlinear and complex. Model predictive control (MPC), as the only one among the so-called advanced control techniques, owns tremendous success in many practical fields in recent decades. Based on the features of determining the capability of the current controller and locating the best achievable performance, control performance assessment (CPA) techniques have received a great interest from both academia and industry. In order to apply CPA technique to the wastewater treatment process, the MPC controller should be designed and tuned first. In this paper, the MPC strategy is introduced to the Benchmark Simulation Model 1 (BSM1). Under this circumstance, the control target is to maintain the nitrate concentrate in the second compartment and the dissolved oxygen concentrate in the fifth compartment simultaneously at their desired values using the MPC control strategy. Three tunings of MPC parameters are compared for MPC design. In this paper, MPC is applied to control dissolved oxygen concentration and nitrate concentration. The improvement potential of the MPC controller is figured out and assessed by using multivariate control performance assessment approach.