Analysis And Design Of A Fuzzy PID Controller Based On T-S Model

The basic PID controller has the merits of simpler algorithm and high reliability.For a certaim controlled plant, it can get a satisfied controlling effect by properlyregulating its three parameters. However, for an uncertain complex controlled system,which has the features of big lag, nonlinear and variable parameters with time, it cannot get an expected result. Fuzzy controller has outstanding merits of no relying oncontrolled plant's mathematic model and of high adaptability, but it has errors in thestable state. Therefore, upon the features of PID controller and Fuzzy controller, bycombining them, a new Fuzzy PID controller can be presented. The new controllercan have both merits for them and can get expecting controlling effect for theuncertain complex controller system and different controlled target.Although the existing Fuzzy PID controllers have their merits, they inevitablyhave some defects. For the purpose of making some improvement, the peaper developa new T-S Model Based Fuzzy PID Controller.First, the defects of the basic PID controller and the fetures of the Fuzzy PIDcontroller are presented as well as the developing trend and several main studyorientation of the Fuzzy PID controller.Second, by analyzing the T-S model, for the purpose of improvement, a new kindof Advanced Fuzzy PID controller has been designed. The main improvements of thenew controller lie in combining the Fuzzy control and T-S model, reducing thenumber of consequent parameters and creating better control performance through theaction of flexible factor in the condition of less fuzzy rules. As a result, the newcontroller is more simple and practicable with specific physical meanings.Finally,we compared the control performance between the T-S Model BasedFuzzy PID controller and basic PID controller through Matlab simulink for differentsystems. The results show us this controller has merits of Fuzzy controller and basicPID controller. It has strong robustness and improves the dynamic and staticproperties of controlled systems.