Otomatisasi Sistem Bendung menggunakan Metode Backpropagation untuk Mengatur Debit Air berbasis Internet of Thing

Abstract

Weir is a controlling and monitoring tool of all water governance and also as an anticipation of flooding. The weir is used to address the amount of water discharge that potentially creates flooding in an area. Flooding occurs with high rainfall, with high bulk resulting in the water height of the weir steadily increasing drastically. The design of the overview was developed into two parts, Namely the Internet of Thing-based monitoring system using smartphones and controlling debit by moving the water door. To move the water door is based on the water height and calculate the discharge of the outgoing water. The controller as the main controller to separate the data to compute the prediction and run the water door opening control algorithm. The architecture of a synthetic neural network 5-8-1 which uses a hidden layer of eight nodes and one output. The Backpropagation method can classify the floodgates with an accuracy value of 91.78% in the training process. In testing results resulted in an error value of 12.98%.