| A 3D Fuzzy PID & CMPC hybrid strategy is presented for advanced control and optimization of nonlinear processes. The strategy utilizes fuzzy logic in a three dimensional fashion (i.e. acting upon error, change in error, and the controlled variable) to adjust the tuning parameters of a PID controller. Fuzzy logic provides a vehicle for incorporating nonlinear process knowledge into the common PID algorithm via tuning parameters that adjust to the process state at each sampling interval. Using a 3D Fuzzy PID controller to handle the process nonlinearity, then it is possible to develop a linearized model between the set point of the controller and the process output. CMPC can be then employed to provide advanced control and optimization. The 3D Fuzzy PID & CMPC hybrid strategy is applied to both a pilot-scale nonlinear system containing pH-type nonlinearity as well as an industrial coal-fired boiler located at the University of Louisville Steam Plant.; For the industrial coal-fired boiler, an experimental 3D fuzzy logic knowledge base is applied to the master pressure PID controller to handle an apparent nonlinearity. [ONLINE]®, a CMPC software package developed by Six Sigma and Advanced Controls, Inc. is implemented under supervision of a Master Fuel Table for real time optimization (RTO). The boiler combustion process is a four-input by three-output system with one measured disturbance. The three controlled variables of interest are steam pressure, stack O 2, and opacity. The four PID controllers that regulate these variables are the master pressure, forced draft, O2 trim, and induced draft. The measured disturbance is the steam demand (flow to campus).; The 3D fuzzy PID & CMPC hybrid strategy demonstrated excellent performance in both the short and long term by holding pressure at set point and providing the optimal mix of fuel and combustion air. During the month of December 2002, a 2-ton per day reduction in coal usage as well as an increase of 5.2% in combustion efficiency was observed. |
