Implementation of neural network based on the microcontroller for energy saving of electric kettle

This paper presents the implementation of neural network based on Arduino Uno R3 microcontroller for energy saving of electric kettle in the office. The neural network uses back propagation technique and levenberg-marquardt algorithm for learning and training. The day and usage time of electric kettle are inputs for learning and training until the neural network can predict and control the operation time correctly when the user needs to use the electric kettle. As the results, the neural network can predict and control the operation time of electric kettle including amount of water usage for each day accurately. Amount of water usage for each day is based on the average value about 760 cubic centimeters. This neural network can also provide the average of energy saving approximately 49.84 percent as comparing with the general electric kettle. This technique may also apply with the other home applications.