Performance evaluation of a self-tuning fuzzy controller

Using a self-tuning technique based on fuzzy metarules, the authors implemented a self-tuning fuzzy controller for a gas-fired water heater. They describe the implementation and its simulation results. The problem was to replace a proportional-integral-derivative (PID) control algorithm with a fuzzy controller, using initial guesses as to the fuzzy membership functions and rules; to tune the fuzzy controller for optimum performance; and to compare the performance results to those from a PID controller. When tuned for a given set of operating conditions, the fuzzy controller matches or exceeds the PID controller performance.<
>