Seoul National University professor Byeng Dong Youn, team, PHM Asia Pacific 2017

PHM Data Challenge Winner (1st and 2nd place)

OnePredict Inc. is a company founded by Professor Byeng Dong Youn of Seoul National University and Professor of Systems Integrity and Risk Management Laboratory, from July 12th to 15th, 2017 in Jeju Hotel, Ramada Plaza. He participated in the PHM Data Challenge hosted by Society 2017 (hereinafter referred to as PHMAP17) and won the first prize and the second prize at the same time. The PHM Data Challenge is an international convention on Prognostics and Health Management (PHM) technology that focuses on teams that have achieved good results in the process of predicting system operation through real-world data . Professor Youn has been selected as a winner in the IEEE Reliability Society in 2014 and the PHM Data Challenge in 2014 and 2015, and this time he has won the world championship We have proved once again that we have the best technology.

In this conference, the problem of predicting the type and timing of the failures of the components constituting the gear box inside the differentiator system of the thermal power plant was raised. Training data consisted of three different types of vibration data that were in normal operation without failure history. Three different types of vibration data were provided by the test data before failure. The SHRMer team, ranked number one, succeeded in predicting accurate failure times using weighted regression techniques after analyzing frequency characteristics and Cepstrum changes over time.In addition, the Starling team, which was ranked 2nd, separated the fault signals using Wavelet Decomposition Packet technology and proposed a method to predict the failure in various conditions by using autocorrelation function.

It is also noteworthy that the theme of the competition was the power plant. Recently, we have already experienced the worst power shortage in the past due to the breakdown of the power plant. In light of the fact that unforeseen power plant failures cause unimaginable losses, the development and dissemination of technologies that can prevent them beforehand is a social responsibility and mission entrusted to the engineering profession. Therefore, the researchers who participated in the PHM Data Challenge welcomed the event as an important turning point for renewed recognition of the maintenance culture of the Korean power plant.

Professor Yoon''s approach and results presented at this conference will be published as an invited paper in international journals on failure prediction and health management.<?xml:namespace prefix = "o" ns = "urn:schemas-microsoft-com:office:office" />