R-based neural networks decision model for water air cooler system

Abstract

Water air cooler is a process that uses an evaporative system. In its first generation, there was no temperature sensor, and its fan was turned by an electric motor in an open loop system with a fixed speed. As well, the water’s flow and level in the tank are governed by a mechanical system, which is generally a floating ball attached to a shaft. In order to ameliorate this classical system with more advantages and performances, many ideas are included in subsequent generations such the integration of embedded systems, intelligent control and components of manufacturing materials. In this paper, the current study aims to integrate an intelligent system, which is the neural networks by using R language to give a smart decision model to command relays switching dedicated to control the electric motors, where the first one is tied up with a fan and the other to an electro-pump. The HC-SR04 ultrasonic and DHT11 sensors supervise the two desired parameters control, water level in the tank and the outside temperature successively.