he speed control of three-phase induction motor (IM) is quite complex due to its nonlinear characteristics. Therefore, controlling the flux and torque parameters with proper decoupling is derived from speed reference feedback. Classical speed control (indirect/direct vector control) of IM drives uses proportional-integral (P-I) and/or proportional-integral-derivative (P-I-D) controllers that have constant gain values at all operating conditions. In addition, the slip calculation relies on rotor time constant, but it varies with operating conditions. These controllers are not adaptive in nature with respect to the operating condition. Neural network and fuzzy logic are said to be intelligent, used to overcome the above drawbacks [1], [2], [3] and [4]. But neural network controllers (NNC) do not involve analytical model of the complete system under test and do not have the ability to adapt it to change in control environment. Still, it is a tedious process to select appropriate neural controller architecture and its training neuron process. Moreover, FL is the simplest of intelligent controller versions and uses expert knowledge to drive the system even if the system is undefined and also with parameter variation issues [5] and [6].