Predictive Maintenance of Auxiliary Converters in Electric Locomotives using Machine Learning and Artificial Intelligence

다국어 초록 (Multilingual Abstract)

This research focuses on the benefits of implementing Predictive Maintenance (PM) methods and how they can be applied to the maintenance of Auxiliary (AUX) converters in several electric locomotive series of Uzbekistan Railways JSC.

목차 (Table of Contents)

• CONTENTS
• 1. Introduction 1
• 1.2 Structure of the research 1
• 1.3 Brief History of Railway in Uzbekistan 2
• 1.4 Modern electric locomotive series in Uzbekistan Railways 3
• 1.5 Challenges in the Maintenance of Current Electric Locomotives in Uzbekistan Railways 9
• 1.6 Concept of Predictive Maintenance 11
• 1.7 The aim of the project and its objectives 13
• 2 AUX Converters and AUX Machines 15
• 2.1 AUX converters of electric locomotives and their functions 15
• 2.2 Root Causes of Failures in AUX Converters and data collection 18
• 3 Research methodology 22
• 4 Fault Analysis of AUX converter faults 24
• 4.1 Data Extraction Methods for Fault Analysis 24
• 4.2 Data Analysis of AUX Converter Faults Across Multiple Sample Sets 24
• 4.3 Data Cleaning Techniques for the fault analysis of AUX Converters 25
• 4.4 Evaluation and Selection of Effective Plotting methods for AUX converter fault analysis 28
• 5 PM and ML application in forecasting AUX converter faults using Python 30
• 5.1 Exploratory Data Analysis (EDA) of the current dataset 31
• 5.2 Identification of most common faults of AUX Converters 35
• 5.3 Review of forecasting models for the selected fault datasets 37
• 5.4 Testing statistical forecasting models for the selected fault datasets 39
• 5.5 Testing ML and DL models for the selected datasets 42
• 6 Results and discussion 45
• 7 Conclusion 47