Adaptive Neuro-Fuzzy Inference System Based Genetic Algorithm Tuned Fractional Order Proportional Integral Derivative Speed Control of Vector Controlled Induction Motor

Abstract

In the industrial sector especially in the area where variable frequency drive is needed, induction motors play a vital role. So without proper controlling of the speed, it is virtually impossible to achieve the desired task for a specific application using an induction motor. Three-phase Induction motor, specifically the squirrel cage induction motors are exhibits a nonlinear behavior on their sudden changes in load and variable speed applications as well as parameter variations. That is why an advanced controller is needed to enhance induction motor performance. Thus, this thesis aims to design an Adaptive Neuro-Fuzzy Inference System based Genetic Algorithm Tuned Fractional Order Proportional Integral Derivative speed controller for a vector-controlled induction motor in which parameters of Fractional Order Proportional Integral Derivative are optimized by using a genetic algorithm. So that the tracking performance of the induction motor drive will be improved. To achieve this, mathematical modeling is done based on vector control, and also for implementation purposes, the corresponding model is done using MATLAB software tools. The performance of induction motor based on the Fractional Order Proportional Integral Derivative speed controller was evaluated under the application of sudden load change, motor parameter variation, and giving variable reference speed. Accordingly, the GA- Fractional Order Proportional Integral Derivative speed controller has a settling time of 0.412sec, a rising time of 0.355, 9% overshoot, and 7% steady-state error under 120 rad/s rated speed. On the other hand, the Adaptive Neuro-Fuzzy Inference System based Genetic Algorithm Fractional Order Proportional Integral Derivative speed controller has better performance than Genetic Algorithm Tuned Fractional Order Proportional Integral Derivative and it has a settling time of 0.403sec, rising time of 0.307sec, 6% overshoot, and 0.5% steady-state error. In general, the performance evaluations of the implemented controllers revealed that Adaptive Neuro-Fuzzy Inference System based Genetic Algorithm Tuned Fractional Order Proportional Integral Derivative speed controller along with space vector pulse width modulation outperformed Genetic Algorithm Tuned Fractional Order Proportional Integral Derivative.