Microcontroller based implementation of a fuzzy knowledge based controller

Abstract

In recent times, fuzzy logic has been used and applied in wide areas, starting from consumer electronics like washing machines to robotics to many industrial control systems like temperature controllers for process plants. Our work describes an implementation of fuzzy logic control algorithm using inexpensive hardware to control the temperature of a system, without any special software tools. A cooling system generally involves complex and time-variant plant, with delays and non- linearity, and often with poorly defined dynamics. Fuzzy logic control algorithm solves problems that are difficult to address with traditional control techniques, and at the same time provides us with a response better than conventional PID controllers. In the present work, this has been proved with the help of MATLAB simulations. Thereafter the program for the fuzzy control algorithm is written in C++ language and implemented through ARDUINO UNO tool kit. Further system functional is tested and the performance is evaluated taking several set-points and disturbances into account. The performance of the hardware is compared with that of MATLAB simulations of the same case and the results are verified.