Automated Smart Home Controller Based on Adaptive Linear Neural Network

Abstract

The purpose of this research is to model a smart home appliances control system using a neural network method. This control system consists of several components i.e. rain sensor, temperature sensor, light sensor, stepper motor and LCD on Arduino. Adaptive linear neural networks called Adaline algorithm is used for the training process of artificial neural networks. The inputs and targets are trained with a network that has been built to get the learning weights to be used as the basis for calculations on the next incoming training data. Adaline algorithm is used for the training process of Artificial Neural Networks. The results of the output of light intensity are 0, 150, 300, 400 with lux units and the results of the intensity of rain are 0, 250, 400, 700 in mm units of each water humidity simulated in light and rain sensors. The input obtained will be processed into an output that will move the motor to the position that has been set. Contribution of the results of this study, a proposed neural network-based multi-control system model based on criteria values of Adaline.