Detection of abnormal engine using the impulsive vibration signal

Hyun-Soo Byun; Sung-Hwan Shin; In-Soo Jung; Jae-Min Jin

doi:10.7776/ASK.2025.44.2.109

All Issue

2025 Vol.44, Issue 2 Preview Page Next Page

Research Article

Detection of abnormal engine using the impulsive vibration signal 임펄스성 진동 신호를 통한 엔진 이상상태 진단

31 March 2025. pp. 109-116

PDF XML

Abstract

The vibration signal generated from the vehicle engine exhibits continuous impulsive characteristics that vary depending on the engine’s condition. Conventional vehicle maintenance involves checking each component individually, but this process is complex and requires a lot of time, cost, and labor. This study aims to detect abnormal engine conditions using impulsive vibration signals. Initially, in the analysis of engine abnormal conditions, quantitative factors that can normalize impulsive signal characteristics regardless of overall level were introduced, and classification according to engine abnormal conditions was conducted. However, some samples presented ambiguity in discrimination. To optimize processing for continuously calculated sequential data and minimize discrimination errors from a statistical perspective, Hidden Markov Model (HMM) method was employed to enhance detection performance. The HMM involves four steps: model design (steps 1 ~ 3) using 70 % of the data and validation with the remaining 30 % of the data. Through HMM analysis, it was confirmed that classification according to abnormal conditions was successfully performed for all samples measured for validation in this study.

Keywords

Abnormal condition

Diagnostics

Impulsive signal

Engine

Hidden Markov Model (HMM)

차량 엔진에서 발생하는 진동 신호는 연속적인 임펄스성 특징을 가지며, 엔진의 상태에 따라 차이가 발생한다. 기존의 차량 정비는 일일이 부품을 체크하여 엔진 정비를 진행하지만, 이러한 과정은 복잡하고 많은 시간과 비용, 노동력을 요구한다. 본 연구는 엔진에서 발생하는 임펄스성 진동 신호를 이용하여 엔진의 이상상태를 검출하고자 한다. 먼저 엔진의 이상상태 분석에는 임펄스성 신호의 특성을 레벨에 관계없이 정량화 가능한 계산인자를 적용하였으며, 엔진 이상상태에 따른 분류가 수행됨을 확인하였다. 하지만 일부 샘플에 대하여 판별의 모호성이 나타나는 문제점이 발생하였다. 따라서 연속적으로 계산된 순차데이터 처리에 최적화되어 있으며, 통계적 관점에서 분석을 실시하여 판별 오류를 최소화할 수 있는 은닉마르코프모델(Hidden Markov Model, HMM) 방법을 도입하여 검출 성능을 향상시켰다. HMM은 4단계로 구성되어 있으며, 70 %의 데이터를 통해 모델 디자인(1~3단계)을 수행하며, 나머지 30 % 데이터를 통해 검증을 수행한다. 결과적으로 HMM 분석을 통해 본 연구 검증을 위해 측정된 모든 샘플에 대하여 이상상태에 따른 분류가 정상적으로 수행됨을 확인하였다.

키워드

이상상태

진단

임펄스성 신호

엔진

은닉마르코프모델

References

Gyeonggi-do transport information center, https://gits.gg.go.kr/web/newboard/5567.do?mode=view&idx=194656682, (Last viewed August 11, 2024).

R. B. GmbH, Diesel-Engine Management, 4th Edition (Wiley, New York, 2006), pp. 60-77.

C. G. Lee, J. H. Lee, H. H. Lee, and D. Y. Lee, "A study on service life restoration of control valve for a small diesel common rail injector" (in Korean), Trans. KSAE. 26, 513-520 (2018).

10.7467/KSAE.2018.26.4.513

O. Asi, "Failure of a diesel engine injector nozzel by cavitation damage," Eng. Fail. Anal. 13, 1126-1133 (2006).

10.1016/j.engfailanal.2005.07.021

I. S. Jung, D. C. Lee, K. J. Noh, and J. H. Chang, "Development of AI-based diagnosis technology using noise/vibration data of automobile powertrain" (in Korean), Trans. Korean Soc. Noise Vib. Eng. 32, 37-44 (2022).

10.5050/KSNVE.2022.32.1.037

C. H. Kim, Dynamic sound design for 4-cylinder engine using modifying its harmonic order, (M.S. thesis, Kookmin University, 2017).

J. C. Kim, H. Ji, K. Lee, K. Moon, and J. Seo, "Analysis of vibration parameters for the fault diagnosis of reduction unit for high-speed train," J. Korean Soc. Precis. Eng. 30, 679-686 (2013).

10.7736/KSPE.2013.30.7.679

S. -H. Shin, S. R. Kim, and Y. -H. Seo, "Development of a fault monitoring technique for wind turbines using a hidden markov model," Sensors, 18, 1790 (2018).

10.3390/s1806179029865235PMC6021971

H. S. Byun, Detection of fault cylinder using the vibration signal, (M.S. thesis, Kookmin University, 2022).

Information

Publisher :The Acoustical Society of Korea
Publisher(Ko) :한국음향학회
Journal Title :The Journal of the Acoustical Society of Korea
Journal Title(Ko) :한국음향학회지
Volume : 44
No :2
Pages :109-116
Received Date : 2025-01-21
Revised Date : 2025-02-13
Accepted Date : 2025-02-17
DOI :https://doi.org/10.7776/ASK.2025.44.2.109

The Journal of the Acoustical Society of KoreaISSN:1225-4428(Print) 2287-3775(Online)한국음향학회

All Issue