You are on page 1of 9

Adaptive Fuzzy

Controller
Submitted by:
Engr. Umer Ehsan SP17-REE-011
Engr. Waleed Iqbal FA17-REE-021
Engr. Rashid Ashraf SP17-REE-021
Controllers
 Responsible for the performance of the control system
 It is a device or an algorithm that works to maintain the value of the
controlled variable at set point
 A control system can control its output(s) to a particular value or
perform a sequence of events or perform an event if the specified
conditions are satisfied based on the input(s) given.
Adaptive Controllers

 A controller which can use the information it gathers online to change itself
and improve its performance
 It "adapts" to the feedback loop it is inside, and develops into a better
controller over time.
 It does not need a priori information about the bounds on the parameters
Types of Adaptive Controllers

 Feed-forward adaptive control


 Feedback adaptive control
Methods:
 Direct method (estimated parameters are directly used)
 Indirect method (estimated parameters are used to calculate required
controller parameters)
 Hybrid method (rely on both estimation of parameters and direct modification
of the control law)
Fuzzy Adaptive controller

 Adaptive Fuzzy Controller is designed with some adjustable parameters along


with an embedded mechanism for adjusting them.
Steps for Designing:
 Collection of observable data - The observable data is collected to calculate
the performance of controller.
 Adjustment of controller parameters - Now with the help of controller
performance, calculation of adjustment of controller parameters would be
done.
 Improvement in performance of controller - In this step, the controller
parameters are adjusted to improve the performance of controller.
Adaptive neuro-fuzzy inference system (ANFIS)

 Artificial neural network that is based on Takagi–Sugeno fuzzy Inference


system
 It integrates both neural networks and fuzzy logic principles
 Its inference system corresponds to a set of fuzzy IF–THEN rules that have
learning capability to approximate nonlinear functions
 Using a given input/output data set, the toolbox function anfis constructs a
fuzzy inference system (FIS) whose membership function
 Allows fuzzy systems to learn from the data it is modeling
Why do we use ANFIS?

 In situations in which we can’t discern what the membership functions should


look like just by looking at the data
 Instead of choosing arbitrary parameters for membership function we tailor it
according to input/output data based on neuro-adaptive learning techniques
 It learns information about the data set in order to compute the membership
function
 Used in plants that are mathematically poorly understood and where the
experienced human operators are available

You might also like