Decentralized PID Controller Tuning Based on Desired Dynamic Equations

Abstract

The paper aims to show the ability of the desired-dynamic-equations (DDE) based decentralized PID controller tuning method with its application to the ALSTOM gasifier benchmark control problem. The DDE-based PID controller tuning method was deduced from a kind of nonlinear adaptive controller with relative degree of two, which behaves good tracking performance and robustness by using an extended state observer and DDE to estimate and compensate the uncertainty and disturbance. The tunable parameters of DDE-based PID controller have explicit and distinctive physical meanings, thus can be tuned separately rather than iteratively as traditional PID controller does. The tuning method is firstly applied to the ALSTOM gasifier benchmark control problem with linear model, simulation shows that it exceeds output limits only twice (the least in literature) at 0% load with the step and sinusoidal disturbance; then the same controller is applied without any modification to nonlinear gasifier model, and the simulations show that it not only meets all control specifications, but also follows load change rapidly and shows good adaptability to the coal quality change.