No abstract

No abstract

There are many design parameters in the structure of fuzzy logic controllers. Conventional methods that don't have a systematic approach are often used in determining of these parameters. However, setting the controller parameters in this way leads to long experiments and this takes a lot of time. For this reason, design parameters of the fuzzy logic controller are usually determined by using heuristic algorithms. Because, heuristic algorithms can offer solutions that are very close to the optimal solution for the problems where exact solution cannot be obtained. In this study, output membership functions of a fuzzy logic controller are optimized using particle swarm optimization and genetic algorithm. Design and optimization stages are explained in detail and results are compared with each other.

Alternative current motors are generally modelled with constant system parameters. However, parameters of the motor show variations during operation due to complex, unidentified and nonlinear system dynamics. Fuzzy logic controllers are widely used as a solution to overcome this problem. Because, fuzzy logic controllers don't be affected from model uncertainties of the system to be controlled. In this study, permanent magnet synchronous motor that is one of the alternative current motors is controlled with fuzzy logic controller. All stages of the designed controllers are developed by using unique softwares instead of ready toolboxes. Triangular and trapezoidal membership functions with Mamdani and Larsen inference methods are used in the designed fuzzy logic controllers. Speed control performances of the fuzzy logic controllers have been investigated and compared with each other in the simulation studies. According to the obtained results from simulation studies, it is observed that fuzzy logic controllers designed with Larsen fuzzy inference method and trapezoidal input membership function provides better performance.