Modeling of fuzzy logic controller with artificial neural network

Fuzzy Logic Controller (FLC) is a fuzzy set-based controller structure that can be formed close to human thinking logic and provides effective solutions in the control of nonlinear systems. It contains many mathematical operations that cause it to react slowly while providing effective results in terms of control performance. Artificial Neural Network (ANN) can be used to obtain models of many systems. The parallel operation feature of ANN makes it possible that it works fast. In addition, the ANN model has a simpler structure than FLC, which makes it easy to implement the controller in terms of hardware and software. In this study, it is aimed to obtain a model of FLC with a short response time by using ANN. FLC and FLC's ANN model were designed with the help of toolboxes in the MATLAB software. The performance of the proposed method has been tested by simulation studies on the control of the inverted pendulum system using the Simulink add-on. A data set was created using the input and output data of FLC and ANN was trained with this data set. Thus, an ANN model has been obtained that can model the FLC very well and give the same output values as FLC to the applied input values. The comparison of FLC and FLC's ANN model were performed by running with a different data set. In the simulation studies carried out in Simulink add-on, it has been observed that the control performances of FLC and FLC's ANN model are very close to each other and the response time of the ANN model is much faster than the response time of the FLC as targeted. The proposed method can be used in control applications that need fast response.