International Conference on Electrical Engineering, Computer Science and Informatics (EECSI), vol: 4, (2017)
Wireless sensor system for prediction of carbon monoxide concentration using fuzzy time series
Suryono S., Surarso B., Saputra R., Bardadi A.
Abstract
Carbon monoxide (CO) concentration produced from incomplete material burning affects both work health and safety. A smart system capable of early detection of carbon monoxide (CO) concentration is therefore required. This research develops a carbon monoxide sensor detection capability using a wireless sensor system that transmits data to the web server via internet connection. A semiconductor CO sensor is installed in a remote terminal unit. A computer application is developed for data acquisition and sending via online and in real time to a web server using an internet modem. For a web-based prediction of CO concentration, a Fuzzy Time Series algorithm induced by Pritpal Sing matrix is used. This research uses CO concentration data for two months. The resulting carbon monoxide concentration prediction is displayed in real time on a dashboard. This prediction is for the next days forecast. Results show that the Fuzzy Time Series that is induced by Pritpal Sing matrix has an average error of 2.67 %, calculated with its average forecasting error rate (AFER). This error value varies, depending on the number of data and data characteristics. © 2018, Institute of Advanced Engineering and Science. All rights reserved.
Keyword: Error percentage; Internet connection; Prediction; Real time; Wireless