Affiliations: Department of Electronic Engineering, GangneungWonju National University, Jukheon-gil, Gangneung, Gangwon-do, South Korea
Correspondence:
[*]
Corresponding author. Moon-Sik Kang, Department of Electronic Engineering, GangneungWonju National University, 7 Jukheon-gil, Gangneung, Gangwon-do, 25457 South Korea. Tel.: +82-33-640-2383; Fax: +82-33-646-0740; E-mail: mskang@gwnu.ac.kr.
Abstract: This paper proposes an intelligent noise prediction system using pattern analysis and deep learning method. The proposed system is designed to analyze an interlayer noise information from noise and vibration sensors in order to provide a noise warning signal to the noise generator. This scheme is based on both a pattern analysis and a predictive model at that time when the threshold is exceeded. The analyzing process of noise information is performed on Arduino board and noise data is collected from multiple sensors with Bluetooth transmission technology. The Firebase cloud server allows users to access the collected data easily. In addition, a large amount of stored data is analyzed systematically using both R program and a predictive model according to a noise pattern. Performance analysis is carried out to evaluate the excellent performance of the proposed system by showing the accurate predictive results.
Keywords: Noise prediction system, deep learning method, pattern analysis, Firebase cloud server, noise and vibration sensors
