Design and Analysis of PID and Fuzzy Control System for Stewart Platform with Genetic Algorithm as Parameters Tuner
PID control system has dominated the industrial field for many years, but it relies heavily on control engineering's expert knowledge to tune the parameters. In this thesis we would like to solve that problem with expert programs. We choose the Stewart Platform parallel robot from MATLAB because it's a good model to study with widely accepted design for a motion control device and an accurate positioning capability. First we simulate the initial Stewart Platform with PID controller and save the results. Then we build a new control system with fuzzy controller using inputs/outputs data set training with ANFIS as initiator. After that we tune the PID parameters (Kp, Ki, Kd) and fuzzy input membership functions using genetic algorithm as the expert system. We compare all the results (the error performance and the controller inputs/outputs responses comparison) and the conclusion are: genetic algorithm very good for PID controller tuning because it has fewer parameters and a solid and clear search range, while GA find difficulties in tuning the fuzzy controller parameters because of the parameters size is to big and uncertainty in search range.
|M00084||(wh)||Available but not for loan - Missing|
No other version available