Control of a coagulation chemical dosing unit for water treatment plants using MMPC based on fuzzy weighting

A multiple model predictive control (MMPC) scheme based on fuzzy weighting is proposed to control the coagulation chemical dosing unit for water treatment plants. The system is represented by a set of linearised models, where each model corresponds to one of the operating regions of the dosing unit. MPC local controller is designed for each local model. The local control input sequence of each model/controller pair in the scheme is aggregated by fuzzy weighting into a global control input to regulate the surface charge and pH level of the dosing unit. The parameters of the fuzzifiers are fine tuned to improve the performance of the proposed control scheme using the nonlinear curve fitting optimisation technique. Simulation results on a process model of the coagulation chemical dosing unit show the ability of the proposed fuzzy weighting MMPC algorithm to transit smoothly from one operating region to another effectively better than switching and weighting MMPC schemes.