A Novel Parameter Estimation Method for First Order Plus Dead Time Loop Using Time Domain Approximate Estimation

To First Order Plus Dead Time (FOPDT) loops, PID control systems with Smith Predictor have more perfect control accuracy and real-time, and its prerequisite is FOPDT loops' parameters are estimated precisely and in real-time. Owing to the sampling characteristics of FOPDT loops, the conventional parameter estimation method, Fuzzy model, cannot calculate their parameters precisely and in real-time to offer an inaccurate and incomplete parameter set of FOPDT loops, so a novel method using Time Domain Approximate Estimation (TDAE) is proposed. At first, the influence that parameter mismatch for FOPDT loops causes to the performance of PID control systems with Smith Predictor is analyzed. Secondly, time domain equations of FOPDT loops are optimized with TDAE by analyzing sampling characteristics, and their parameters are described precisely with temperature using statistical method and Radial Basis Function Neural Network (RBFNN) to form TDAE model. Then, TDAE model and Fuzzy model are tested in heating and cooling experiments, and their accuracy and real-time are compared. The experimental results show that the overshoots of PID control systems with Smith Predictor are reduced and the adjustment time is improved, which means TDAE model is more accurate to estimate FOPDT loops' parameters to enhance the performances of PID control system obviously. Moreover, TDAE model can be used in all kinds of FOPDT loops as a universal methodology.