Study On The Controller Tuning Based On Fuzzy Theory

In this paper, the generation, development, forecast and the characteristic of Fuzzy control were introduced. Advantage and deficiency of Fuzzy control were summed up. Research on the present situation on Fuzzy control was introduced. Foundation of fuzzy control algorithm was introduced in detail. The development and application of traditional PID were studied. An outline of the operation principle of PID controller was given. And the insufficient aspect of PID controller was pointed out. Combined with the fuzzy control and the PID control, Fuzzy PI control and T-S model based on Fuzzy PID control were proposed. The basic principles of IMC control was introduced, combined with IMC control and fuzzy control, a Fuzzy IMC control that on-line tuning of the filter parameters X was given.With the help of Simulink and Fuzzy toolbox in Matlab, Fuzzy PI control, T-S model based fuzzy PID control, as well as the IMC control which tuning the filter parameter online were used to simulate on the resistance furnace system respectively.Through the step response curve of the resistance furnace system analysis, in accordance with the change trends of error and change in error, the corresponding fuzzy rule was established. An incentive factor was introduced based on the dynamic characteristics of the ideal response joined. The incentive factor was determined according to the change tendency of the error and change ratio in error as well as the ideal dynamic response characteristic.First, a two-inputs-three-outputs Fuzzy PI controller was designed, in which the error and change ratio in error were taken as the inputs, andΔK_p,ΔK_i incentive factors X were taken as the outputs. The controleffect of the Fuzzy PI without incentive factor control was improved compared with the PID control.Then a T-S model based Fuzzy PID controller with two inputs was designed. Take the error and change ratio of the error as the inputs and the consequent parameters as the output. An incentive function f(x) was used in control action. The control effect of T-S based Fuzzy PID control without incentive function was enhanced in turn compared with the PID control.Finally aiming at the uncertainty of modeling precision in industrial processes, one kind of Fuzzy IMC control was designed taken the error and change ratio in error as two inputs and the filter parameter as the output. Fuzzy controller with incentive factor was studied as well. Control effects were given under conditions of model mismatch.