Optimization of Generalized Predictive Control (GPC) Tuning Parameters By Response Surface Methodology (RSM)
Response Surface Methodology(RSM) was successfully applied to a process simulator for optimization of Generalized Predictive Control (GPC) tuning parameters. Wireless experimental input/output data obtained from process simulator. GPC algorithm which is written in MATLAB is utilized to wireless temperature control experiments achieved by using MATLAB/Simulink program.The efficiency of the GPC is observed by calculating the integral of the square of the error (ISE) and the integral of the absolute value of the error (IAE) from experimental results which was optimized by the application of RSM. The three independent variables, which had been found the most effective variables on the GPC by screening experiments, were determined as NU, N2 and λ as minimum prediction horizon, maximum prediction horizon and control weighting, respectively. The quadratic models were developed through RSM in terms of related independent variables to describe the ISE and IAE as the two response. Based on statistic analysis, optimum GPC tuning parameters ofNU (X1), N2 (X2) and λ (X3) for minimize the ISE were determined to be 1.7922, 1.9453 and 0.0642 and for minimize the IAE were determined to be 1.8880, 1.9752 and 0.0612, respectively. Calculated optimum points of GPC tuning parameters are close to based on ISE and IAE results. The data evaluated from the quadratic model were good agreement with those measured experimentally. The wireless temperature control is successfully applied to the process simulator and wireless control technique is proposed for various application areas.