A Comparative Study of Performance Assessment and Fault Diagnosis Approaches for Reciprocating Electromechanical Mechanism
| Main Author: | |
|---|---|
| Language: | English |
| Published: |
University of Cincinnati / OhioLINK
2016
|
| Subjects: | |
| Online Access: | http://rave.ohiolink.edu/etdc/view?acc_num=ucin1468512813 |
| id |
ndltd-OhioLink-oai-etd.ohiolink.edu-ucin1468512813 |
|---|---|
| record_format |
oai_dc |
| spelling |
ndltd-OhioLink-oai-etd.ohiolink.edu-ucin14685128132021-08-03T06:37:35Z A Comparative Study of Performance Assessment and Fault Diagnosis Approaches for Reciprocating Electromechanical Mechanism Shi, Zhe Mechanics Mechanical Engineering Prognostics and health management railway point machine reciprocating machine Incipient fault detection Auto associative model Fault diagnosis Reciprocating electromechanical mechanism is commonly used for industrial applications, including reciprocating engines & compressor, reciprocating valve, railway point machine, elevator motor etc. The research work in this thesis mainly focus on a typical reciprocating electromechanical mechanism railway point machine performance assessment and fault diagnosis. Point machine switches track between two alternative routes which requires to operate with a high level of safety and reliability and a failure asset has a significantly high chance to cause system delay and fatal accident. With the growing needs for intelligent operation and safety assurance, developing a system for performance assessment, fault diagnosis as well as remaining useful life prediction for such critical assets has gained greater attention. Prognostics and Health Management (PHM) methodologies and techniques are considered to be the foremost enabling technology for achieving those tasks and extensive research has been carried out in the area of PHM system development for point machines. However, there still exists many unmet needs for the current system and studies. For instance, few studies have been investigated based on the low cost onboard signals, which is necessary to monitor widely distributed infrastructures. Also, far too little attention has been paid to the incipient faults which can also lead to catastrophic accidents. Most of the systems are developed based on expert knowledge and diagnostic expertise, which are problem-specific and difficult to apply for other similar assets.The main objective of this study was to develop performance assessment and fault diagnosis approaches based on low cost onboard signals for point machines to detect incipient failure and diagnose faults and to provide a comparative study about the effectiveness of each approach. Feature based and auto-associative residual based approaches were developed and selected PHM techniques were applied and evaluated for both approaches. Feature based approach follows conventional PHM system development methodology which consists of signal pre-processing, data segmentation, feature extraction & selection, health assessment and fault diagnosis. The limitation of feature based approach is that it needs expert knowledge for data segmentation, diagnostic expertise for feature extraction and it is usually problem-specific. The newly developed auto-associative residual (AAR) based approach provides a way to build the degradation assessment model based on residual vector from auto-associative model input and output, which requires minimal prior knowledge of the system and reduces manual processing work for feature extraction and selection. Unsupervised degradation assessment methods were used for performance assessment and selected multiclass classification algorithms were applied for fault diagnosis in this study. The proposed approaches were applied to an Alstom MET F-BANE point machine for validation. Both approaches achieve over 99% fault detection rate and over 95% overall fault diagnosis accuracy for 17 different health conditions. 2016-09-12 English text University of Cincinnati / OhioLINK http://rave.ohiolink.edu/etdc/view?acc_num=ucin1468512813 http://rave.ohiolink.edu/etdc/view?acc_num=ucin1468512813 unrestricted This thesis or dissertation is protected by copyright: all rights reserved. It may not be copied or redistributed beyond the terms of applicable copyright laws. |
| collection |
NDLTD |
| language |
English |
| sources |
NDLTD |
| topic |
Mechanics Mechanical Engineering Prognostics and health management railway point machine reciprocating machine Incipient fault detection Auto associative model Fault diagnosis |
| spellingShingle |
Mechanics Mechanical Engineering Prognostics and health management railway point machine reciprocating machine Incipient fault detection Auto associative model Fault diagnosis Shi, Zhe A Comparative Study of Performance Assessment and Fault Diagnosis Approaches for Reciprocating Electromechanical Mechanism |
| author |
Shi, Zhe |
| author_facet |
Shi, Zhe |
| author_sort |
Shi, Zhe |
| title |
A Comparative Study of Performance Assessment and Fault Diagnosis Approaches for Reciprocating Electromechanical Mechanism |
| title_short |
A Comparative Study of Performance Assessment and Fault Diagnosis Approaches for Reciprocating Electromechanical Mechanism |
| title_full |
A Comparative Study of Performance Assessment and Fault Diagnosis Approaches for Reciprocating Electromechanical Mechanism |
| title_fullStr |
A Comparative Study of Performance Assessment and Fault Diagnosis Approaches for Reciprocating Electromechanical Mechanism |
| title_full_unstemmed |
A Comparative Study of Performance Assessment and Fault Diagnosis Approaches for Reciprocating Electromechanical Mechanism |
| title_sort |
comparative study of performance assessment and fault diagnosis approaches for reciprocating electromechanical mechanism |
| publisher |
University of Cincinnati / OhioLINK |
| publishDate |
2016 |
| url |
http://rave.ohiolink.edu/etdc/view?acc_num=ucin1468512813 |
| work_keys_str_mv |
AT shizhe acomparativestudyofperformanceassessmentandfaultdiagnosisapproachesforreciprocatingelectromechanicalmechanism AT shizhe comparativestudyofperformanceassessmentandfaultdiagnosisapproachesforreciprocatingelectromechanicalmechanism |
| _version_ |
1719440543674531840 |