Manipulating state variables in a fuzzy controller

Tuning a TSK fuzzy control system involves the process of defining a set of polynomial coefficients and then finding appropriate values for those coefficients that provide for an adequate response to a set of global inputs and initial states that the control system may experience. This process of searching for the optimum set of coefficients, or gains, can prove to be a difficult process. A method that can be utilized to simplify this process involves manipulating the state variables of the polynomial equations, as well as the coefficients. Certain non-linearities, found in most every control system, lend themselves to this type of compensation. Amplifier saturation, in a motor control application, is used as an example to show the implementation of this method. The resulting control system is shown to be equivalent to a traditional TSK solution, but with a simplified tuning process.