Wastewater treatment process (WWTP) is difficult to control because its biological, physical and chemical phenomena are complex and they are interrelated and highly nonlinear.
Recently, model predictive control (MPC), which employs a prediction model of the plant to optimize future plant behavior, has been a popular approach for WWTP. Since WWTPs are multivariable, multi-objective optimization of MPC is required.
Kim Kyong Jin, a researcher at the Faculty of Automation Engineering, has proposed a nonlinear multi-objective MPC (NMMPC) to realize a multivariable control for WWTPs.
The proposed multi-objective optimal control comprises a self-organizing radial basis function neural network (SORBFNN) identifier, a model predictive controller and a multi-objective optimization method. He developed the SORBFNN as a model identifier for approximating the online states of dynamic systems. The solution of the multi-objective optimization is obtained by a gradient method which can shorten the solving time of optimal control problems.
The experiments have revealed that the proposed control technique gives satisfactory tracking and disturbance rejection performance for WWTPs.
© 2021 Kim Chaek University of Technology