Adaptive Fuzzy Controller

Spread the love

Adaptive Fuzzy Controller

On this chapter, we’ll focus on what’s an Adaptive Fuzzy Controller and the way it works. Adaptive Fuzzy Controller is designed with some adjustable parameters together with an embedded mechanism for adjusting them. Adaptive controller has been used for bettering the efficiency of controller.

Primary Steps for Implementing Adaptive Algorithm

Allow us to now focus on the essential steps for implementing adaptive algorithm.

  • Assortment of observable information − The observable information is collected to calculate the efficiency of controller.
  • Adjustment of controller parameters − Now with the assistance of controller efficiency, calculation of adjustment of controller parameters could be carried out.
  • Enchancment in efficiency of controller − On this step, the controller parameters are adjusted to enhance the efficiency of controller.

Operational Ideas

Design of a controller relies on an assumed mathematical mannequin that resembles an actual system. The error between precise system and its mathematical illustration is calculated and whether it is comparatively insignificant than the mannequin is assumed to work successfully.

A threshold fixed that units a boundary for the effectiveness of a controller, additionally exists. The management enter is fed into each the actual system and mathematical mannequin. Right here, assume x(t)x(t) is the output of the actual system and y(t)y(t) is the output of the mathematical mannequin. Then the error ϵ(t)ϵ(t) may be calculated as follows −


Right here, xx desired is the output we would like from the system and μ(t)μ(t) is the output coming from controller and going to each actual in addition to mathematical mannequin.

The next diagram exhibits how the error operate is tracked between output of an actual system and Mathematical mannequin −

Mathematical Model

Parameterization of System

A fuzzy controller the design of which relies on the fuzzy mathematical mannequin could have the next type of fuzzy guidelines −

Rule 1 − IF x1(tn)X11AND...ANDxi(tn)X1ix1(tn)∈X11AND…ANDxi(tn)∈X1i

THEN μ1(tn)=Ok11x1(tn)+Ok12x2(tn)+...+Ok1ixi(tn)μ1(tn)=Ok11x1(tn)+Ok12x2(tn)+…+K1ixi(tn)

Rule 2 − IF x1(tn)X21AND...ANDxi(tn)X2ix1(tn)∈X21AND…ANDxi(tn)∈X2i

THEN μ2(tn)=Ok21x1(tn)+Ok22x2(tn)+...+Ok2ixi(tn)μ2(tn)=Ok21x1(tn)+Ok22x2(tn)+…+K2ixi(tn)




Rule j − IF x1(tn)Xok1AND...ANDxi(tn)Xokix1(tn)∈Xk1AND…ANDxi(tn)∈Xki

THEN μj(tn)=Okj1x1(tn)+Okj2x2(tn)+...+Okjixi(tn)μj(tn)=Kj1x1(tn)+Kj2x2(tn)+…+Kjixi(tn)

The above set of parameters characterizes the controller.

Mechanism Adjustment

The controller parameters are adjusted to enhance the efficiency of controller. The method of calculating the adjustment to the parameters is the adjusting mechanism.

Mathematically, let θ(n)θ(n) be a set of parameters to be adjusted at time t=tnt=tn. The adjustment may be the recalculation of the parameters,


Right here DnDn is the info collected at time t=tnt=tn.

Now this formulation is reformulated by the replace of the parameter set primarily based on its earlier worth as,


Parameters for choosing an Adaptive Fuzzy Controller

The next parameters must be thought of for choosing an adaptive fuzzy controller −

  • Can the system be approximated solely by a fuzzy mannequin?
  • If a system may be approximated solely by a fuzzy mannequin, are the parameters of this fuzzy mannequin available or should they be decided on-line?
  • If a system can’t be approximated solely by a fuzzy mannequin, can it’s approximated piecewise by a set of fuzzy mannequin?
  • If a system may be approximated by a set of fuzzy fashions, are these fashions having the identical format with totally different parameters or are they having totally different codecs?
  • If a system may be approximated by a set of fuzzy fashions having the identical format, every with a distinct set of parameters, are these parameter units available or should they be decided on-line?