A fuzzy-evaluated gain scheduling approach for the PID control of a pressure plant

A fuzzy gain scheduling routine is combined with the well-known PID algorithm to allow the new structure to deal with non-linearities of the process. It is shown that both process gain and integral time change along the operation range, due to non-linear behavior of the process. In current gain scheduling applications, the gain and integral time commutation is abrupt, producing internal disturbances to the closed loop. The issue of these variations and their implication on the controller performance is discussed and solved. Fuzzy modeling is used here as an useful tool for solving the commutation problem. Simulation studies and experimental data show satisfactorily how the proposed algorithm improves the performance of the controller over a large operation range.