Abstract
This research proposes the design, simulation and implementation of the optimization of type-2 membership functions for the Average Approximation of an Interval of Type-2 Fuzzy Logic Controller (AT2-FLC) using bio-inspired algorithms, such as Particle Swarm Optimization (PSO). The optimization only considers certain points of the membership functions, the fuzzy rules are not modified, so that the algorithm minimizes the runtime. Based on the concept of swarm intelligence, PSO is applied to membership functions parameter optimization of the AT2-FLC. Implementations and simulations are carried out on the FPGA device using the Xilinx System Generator. The optimization method was coded in Matlab. Comparisons were made between simulation and implementation of the AT2-FLC, to regulate the velocity of a DC motor. We compared the results of the AT2-FLC under uncertainty and the results are discussed. Experiments were performed by changing the number of bits for encoding the AT2-FLC in VHDL.
The main contribution of this research is the design, simulation and implementation of PSO of the AT2-FLC for real applications in FPGA. The AT2-FLC is targeted to a Xilinx Spartan 3AN XC3S700A device using Xilinx Foundation Environment.