Title Page
Abstract
Contents
CHAPTER 1. INTRODUCTION 9
1.1. Research Background and Objectives 9
1.2. Research Scope and Methodology 12
CHAPTER 2. LITERATURE REVIEW 15
2.1. Construction Productivity Estimation Methods 15
2.1.1. Construction Standard Productivity Rate (CSPR) Approach 15
2.1.2. Discrete Event Simulation (DES) Approach 16
2.1.3. Machine Learning (ML) Approach 18
2.2. Optimization Method for Fleet Selection in Earthmoving 21
2.2.1. Non-Heuristic Approach 21
2.2.2. Heuristic Approach 23
2.3. Limitations of Recent Studies 25
CHAPTER 3. DES MODEL AND DATA GENERATION 27
3.1. DES Model Development 27
3.1.1. Site Observations 29
3.1.2. Data Collection 31
3.1.3. Simulation and Analysis 37
3.2. Database Development 39
3.2.1. Sensitivity Analysis and Data Generation 39
3.2.2. Data Preparation 42
CHAPTER 4. MACHINE LEARNING AND MULTI-OBJECTIVE OPTIMIZATION 45
4.1. Machine Learning Methods 45
4.1.1. Neural Networks 47
4.1.2. XGBOOST 49
4.1.3. Random Forest 51
4.2. Evaluation Metrics 53
4.3 Multi-objective Optimization 54
4.3.1. Multi-objective Optimization Problem Definition 55
4.3.2. NSGA-II Method 55
CHAPTER 5. METHOD IMPLEMENTATION & RESULT 58
5.1. ML Model Development 58
5.1.1. Model Training 58
5.1.2. Evaluation and Comparison 61
5.2. Method Implementation 63
5.2.1. NSGA-II Framework 63
5.2.2. Case Study Analysis 67
CHAPTER 6. CONCLUSION 70
6.1. Research Summary 70
6.2. Research Significant and Contribution 72
6.3. Limitations and Future Work 73
REFERENCES 74