Design and Implementation of Neuro-Fuzzy Algorithm for Fractional Order Processes

Abstract

In this project, design and implementation of Neuro-Fuzzy Algorithm for fractional Order Controller System has been carried out. First the simulation study for the fractional order transfer function using different control algorithm has been done. Fractional order system was controlled by ANFIS and PID controller. Comparative analysis between ANFIS and PID has been carried out in this project. From this analysis it is observed that ANFIS gives better performance. The fractional order PI algorithm has been tested for the sensor less speed control of 3-phase induction motor. A comparison has been made for the fractional order PI algorithm against the integer order PI algorithm.

Model reference adaptive control of fractional order system using ANFIS and CANFIS has been discussed. ANFIS (Adaptive Neuro Fuzzy Inference System) has been used for the MISO (Multi input single output) system and CANFIS (Coactive Neuro Fuzzy Inference System) is used for MIMO (Multi input Multi output) system. In this paper fractional order MIT rules are used to generate adaption law for MRAC. Tuning for the ANFIS and the CANFIS has been carried out by the fractional order MRAC tuning rule. Fractional order MIT rule has been used to control fractional order system. The comparative analysis between CANFIS and ANFIS has been shown in this paper. It has been shown that the fractional order MRAC can be replaced by an appropriate neural network either in form of ANFIS or CANFIS.