컨테이너 크레인을 위한 RCGA 기반의 퍼지제어기

Abstract

This thesis presents the RCGA-based fuzzy controller for container cranes which effectively performs set-point tracking control of trolley and anti-swaying control under system parameter and disturbance changes.

The first part of this thesis focuses on the derivation of the mathematical equation and Takagi-Sugeno(T-S) fuzzy model of a nonlinear container crane system. The T-S fuzzy model is described by several fuzzy IF-THEN rules which locally represent linear input-output relations of the system according to operation conditions and their parameters. The fuzzy membership functions are adjusted by a RCGA.

The second part of this thesis presents a design methodology of the RCGA-based fuzzy controller which guarantees the robustness for changes to system parameters and disturbances, the fuzzy state observer which solves the problems of unmeasurable state variables. Sub-controllers are designed using another RCGA, which satisfy the given constraints for each subsystem and then the overall controller is performed with the combination of these sub-controllers by fuzzy IF-THEN rules. The fuzzy state observer is defined from the set of fuzzy rules with the state observer designed using a RCGA for each subsystem in order to solve the estimation error.

The last part of this thesis performs a simulation to demonstrate the efficacy of the proposed methods. In the results of simulation, the fuzzy model with the membership functions adjusted by a RCGA showed almost similar dynamic characteristics compared to the outputs of the container crane for the input signal of step and sinusoidal types. The simulation results for the RCGA-based fuzzy controller showed not only the fast settling time compared to that of LQ controller for the significant change in parameters, reference input, initial conditions, and disturbances, but also stable and robust control performances without any steady-state error. Also, the fuzzy controller with fuzzy state observer demonstrated more robust control performance than that of LQ controller and showed almost similar response characteristics compared to the RCGA-based fuzzy controller.