A Hardware Implementation of SEPIC Converter Using ANFIS for Water Flow Control

Abstract

Water resources are one of the needs for the continuation of life. The majority of people in Indonesia still use a manual water tank filling system, so it requires regular checks. Filling the water tank manually is ineffective because it requires turning on and off the pump first so the use of electrical energy is increasingly wasteful. In this research, a water flow control system was developed to fill water tanks, making it easier to use water both at home and in public facilities. This water flow control system uses a Single Ended Primary Inductor Converter (SEPIC) Converter as a DC motor pump supply voltage regulator. The SEPIC converter has the advantage of being able to increase and decrease the voltage value without changing the polarity. The duty cycle settings on the SEPIC converter are regulated using the Adaptive Neuro-Fuzzy Inference System (ANFIS) control method. The water flow control settings in this system use feedback from the water flow sensor by utilizing the hall effect. Based on simulations and implementation that have been carried out using MATLAB as well as hardware testing, at a set point value of 1.5 liters/minute the output voltage value is 3.9 V, and at a set point value of 3 liters/minute the output voltage value is 6.8 V. Implementation of ANFIS control has proven to be a solution for regulating voltage and has worked more efficiently on loads.