Adaptive Neuro Fuzzy Inference Based Direct Torque Control Strategy for Robust Speed Control of Induction Motor under Highly Variable Load Conditions

Among the most practical speed regulation techniques, direct torque control (DTC) is one of the simplest and excellent speed regulation strategy for induction motors via controlling of torque. The basic idea is to control, the torque and flux. Direct torque control is the first technology used for controlling the motor control variables of torque and flux. This methodology created the motor additional correct and quick torque control, high dynamic speed response and simple to control. The reference value can be calculated using the flux and torque estimated and also motor parameter. In the conventional DTC strategy, speed management of motor is performed by using PI controllers. Usually the performance of conventional PI controller, used in DTC, is found satisfactory throughout with this controller during the motor speed regulation. But the scenario of speed regulation with this controller, extremely enthusiastic about the system load. It is often found that, practically the performance of conventional PI controller based DTC strategy for the speed regulation of induction motor under highly variable load conditions is very poor. To overcome this difficulty, this paper proposed an Adaptive Neuro Fuzzy (ANFIS) controller based DTC speed regulation strategy for efficiently regulate the speed of induction motor in both the transient and steady state portions under highly variable load conditions. The basic idea is to generate an improved control performance by replacing the conventional PI controller with advance ANFIS controller. The performance of proposed controller and traditional PI controller has been incontestable by simulations using MATLAB/SIMULINK 2012 (b). After the complete comparative analysis among the proposed ANFIS based DTC and conventional PI based DTC speed regulation of induction motor, it is shown that, the speed regulation capability of the proposed technique is very robust and efficient as compare to standard one below numerous load variations. In addition to this it is also found that, the proposed speed regulation strategy not only provides regulated speed in the transient and steady state, but also takes 76.0 % less settling time as compare to conventional PI controller based DTC technique.