표제지
요지
Abstract
목차
제1장 서론 18
1.1. 연구배경 및 목적 18
1.2. 연구방법 및 범위 23
제2장 문헌분석 25
2.1. 고속도로의 리모델링 보수공사 25
2.2. 리모델링 시급지수(RMI) 모형 34
2.3. 다중회귀분석과 인공신경망 44
제3장 포장 공용성 예측모형 50
3.1. 자료수집 51
3.1.1. 영향인자 선정 51
3.1.2. 영향인자 수집 54
3.2. 포장 하부물성 예측모형 개발 59
3.2.1. 후보모형 개발 59
3.2.2. 모형 선정 및 적용 71
3.3. 포장 공용성 연 변화량 예측모형 개발 81
3.3.1. 영향인자 전처리 81
3.3.2. 상관성 분석을 통한 변수 선정 90
3.3.3. 후보모형 개발 100
3.3.4. 모형 선정 및 적용 105
제4장 RMIa 모형 개발 120
4.1. 리모델링 시급도 평가 121
4.1.1. 대표구간 선정 121
4.1.2. 주행영상 촬영 및 패널레이팅 130
4.2. RMIa 모형 개발 135
4.2.1. 현재 RMIa 예측모형 개발 135
4.2.2. 현재 RMIa 예측모형 검토 147
4.2.3. 장래 RMIa 예측모형 개발 및 적용 157
제5장 리모델링 우선순위 구간 선정 164
5.1. 리모델링 우선순위 구간 선정방법 165
5.2. 리모델링 우선순위 구간 선정 및 결과분석 171
제6장 결론 176
참고문헌 179
부록 I. 대표구간 선정을 위한 자료의 심각도 분포 185
Table 2.1. Crack Distress of Asphalt Pavement 26
Table 2.2. Deformation Distress of Asphalt Pavement 29
Table 2.3. Loss Distress of Asphalt Pavement 30
Table 2.4. Difference of Short-Term Blockade Repair and Long-Term Blockade Repair 33
Table 3.1. Influence Factor of Change in IRI, RD, and SD 52
Table 3.2. Correlation Coefficient between IRI, RD, SD change and IRIc, SDc 53
Table 3.3. Collection Method of Influence Factors 55
Table 3.4. Skewness and Kurtosis of Raw Data and Standardized Data 63
Table 3.5. Average and Standard Deviation of Variables for Rescaling 64
Table 3.6. Skewness and Kurtosis of IRI, RD, SD Change in Year 82
Table 3.7. The Determination Method of Correlation with Factors and Annual IRI Change 91
Table 3.8. Correlation Coefficient Result with Factors and Annual IRI Change in Normal Asphalt Pavement 92
Table 3.9. Correlation Coefficient Result with Factors and Annual IRI Change in Composite Asphalt Pavement 93
Table 3.10. The Determination Method of Correlation with Factors and Annual RD Change 94
Table 3.11. Correlation Coefficient Result with Factors and Annual RD Change in Normal Asphalt Pavement 95
Table 3.12. Correlation Coefficient Result with Factors and Annual RD Change in Composite Asphalt Pavement 96
Table 3.13. The Determination Method of Correlation with Factors and Annual SD Change 97
Table 3.14. Correlation Coefficient Result with Factors and Annual SD Change in Normal Asphalt Pavement 98
Table 3.15. Correlation Coefficient Result with Factors and Annual SD Change in Composite Asphalt Pavement 99
Table 3.16. Combination Conditions of Candidate Variables for each Factor Group 101
Table 3.17. Number of Candidate Models by Possible Combinations of Candidate Variables 102
Table 3.18. The Sign of Parameter for Each Annual Change Model 105
Table 3.19. The Determination Coefficient of the Candidate Prediction Model and the Final Selected Prediction Model 107
Table 3.20. Independent Variable of Final Selected Model 109
Table 4.1. Severity Threshold of Pavement Performance and Factors 124
Table 4.2. Severity Combination Distribution in the Candidate Representative Sections on Normal Asphalt Pavement 126
Table 4.3. Severity Combination Distribution in the Candidate Representative Sections on Composite Asphalt Pavement 127
Table 4.4. Severity Distribution of Factors in the Selected Representative Sections 129
Table 4.5. RMI Range according to Remodeling Urgency 130
Table 4.6. Example of RMI Panelrating 134
Table 5.1. Selection Result of Remodeling Priority Section 172
Table 5.2. Remodeling Priority Section of Each Route 173
Table 5.3. Existing Remodeling Work Design Section 174
Table 5.4. Existing Remodeling Work Design Section and Selected Remodeling Priority Section according to Study 175
Fig. 1.1. Length and Ratio of Asphalt and Concrete Pavement on Highway 20
Fig. 1.2. Length of Old Asphalt Pavement on Highway 21
Fig. 1.3. Ratio of Normal Asphalt Pavement and Composite Asphalt Pavement on Highway 22
Fig. 2.1. Repair Method of Asphalt Pavement 32
Fig. 2.2. Concept of IRI 38
Fig. 2.3. Concept of SD 39
Fig. 2.4. Concept of RD 40
Fig. 2.5. Error of Multiple Linear Regression Model 45
Fig. 2.6. Concept of ANN 46
Fig. 2.7. ANN Learning Process 47
Fig. 2.8. ReLU Function 48
Fig. 3.1. Calculation Method of ITS and FT 60
Fig. 3.2. Difference of Distribution between Non-standardized Data and Standardized Data 62
Fig. 3.3. Scale of Raw and Rescaled Data for ITS Prediction Model Development 65
Fig. 3.4. Scale of Raw and Rescaled Data for Fracture Toughness Prediction Model Development 67
Fig. 3.5. ANN structure of ITS and FT prediction models according to the number of hidden layers 70
Fig. 3.6. Minimum Mean Square Error of ITS Prediction Representative Models 71
Fig. 3.7. Minimum Mean Square Error of Fracture Toughness Prediction Representative Models 72
Fig. 3.8. Mean Square Error every Learning Epoch of Train and Validation Dataset of Selected ITS Prediction Model 73
Fig. 3.9. Mean Square Error every Learning Epoch of Train and Validation Dataset of Selected Fracture Toughness Prediction Model 74
Fig. 3.10. Scatter Plot of Train Dataset and Test Dataset of Actual and Predicted ITS 75
Fig. 3.11. Scatter Plot of Train Dataset and Test Dataset of Actual and Predicted Fracture Toughness 77
Fig. 3.12. Distribution of Predicted ITS Applied across the Entire Highway 79
Fig. 3.13. Distribution of Predicted Fracture Toughness Applied across the Entire Highway 80
Fig. 3.14. Difference of Entire Data Scatter Plot and Grouped Data Scatter Plot 83
Fig. 3.15. Scale of Raw and Rescaled Data for Independent Variable 85
Fig. 3.16. Correlation of Dependent Variable and Independent Variable to which Various Functions are Applied 88
Fig. 3.17. Determination Coefficient of Candidate Models by Percentage Order 104
Fig. 3.18. Scatter Plots of the Annual IRI Change Model Train Set and Test Set in Normal Asphalt Pavement 112
Fig. 3.19. Scatter Plots of the Annual IRI Change Model Train Set and Test Set in Composite Asphalt Pavement 113
Fig. 3.20. Scatter Plots of the Annual RD Change Model Train Set and Test Set in Normal Asphalt Pavement 115
Fig. 3.21. Scatter Plots of the Annual RD Change Model Train Set and Test Set in Composite Asphalt Pavement 116
Fig. 3.22. Scatter Plots of the Annual SD Change Model Train Set and Test Set in Normal Asphalt Pavement 118
Fig. 3.23. Scatter Plots of the Annual SD Change Model Train Set and Test Set in Composite Asphalt Pavement 119
Fig. 4.1. Representative Section Selection Process 122
Fig. 4.2. Determination Method for Tertile 123
Fig. 4.3. Selection Method for Candidate Representative Section 125
Fig. 4.4. Selection Method for Representative Section 128
Fig. 4.5. Location Distribution of Representative Section 129
Fig. 4.6. Pavement Condition according to RMI Range 131
Fig. 4.7. RMI Panelrating Method of Driving Video of Representative Section 132
Fig. 4.8. Example of Driving Video for Representative Sections 133
Fig. 4.9. Correlation of Panelrated RMI and IRI, log(IRI+1) on Normal Asphalt Pavement 136
Fig. 4.10. Correlation of Panelrated RMI and RD, log(RD+1) on Normal Asphalt Pavement 138
Fig. 4.11. Correlation of Panelrated RMI and SD, log(SD+1) on Normal Asphalt Pavement 139
Fig. 4.12. Correlation of Panelrated RMI and IRI, log(IRI+1) on Composite Asphalt Pavement 141
Fig. 4.13. Correlation of Panelrated RMI and RD, log(RD+1) on Composite Asphalt Pavement 142
Fig. 4.14. Correlation of Panelrated RMI and SD, log(SD+1) on Composite Asphalt Pavement 144
Fig. 4.15. Scatter Plot of Panelrated RMI and Predicted RMI 146
Fig. 4.16. The Relationship between Predicted RMI and IRI for All Sections of Highway Asphalt Pavement 148
Fig. 4.17. The Relationship between Predicted RMI and RD for All Sections of Highway Asphalt Pavement 149
Fig. 4.18. The Relationship between Predicted RMI and SD for All Sections of Highway Asphalt Pavement 151
Fig. 4.19. Sensitivity of Independent Variables in the RMI Model of Normal Asphalt Pavement 154
Fig. 4.20. Sensitivity of Independent Variables in the RMI Model of Composite Asphalt Pavement 156
Fig. 4.21. Scatter Plot of Predicted and Actual Values of IRI, RD, SD and RMI after 2 years 160
Fig. 4.22. Distribution of Predicted and Actual Values of IRI, RD, SD and RMI after 3 years 163
Fig. 5.1. 100m Length Unit Section and Homogeneous Section 164
Fig. 5.2. Concept of IC/JCT Unit Section 166
Fig. 5.3. Process of 1st Homogeneous Section Selection Method 166
Fig. 5.4. Visualization of 1st Homogeneous Section Selection Method 167
Fig. 5.5. Process of Final Homogeneous Section Selection Method 169
Fig. 5.6. Visualization of Final Homogeneous Section Selection Method 169
Fig. I.1. Severity Distribution of Data on Normal Asphalt Pavement 187
Fig. I.2. Severity Distribution of Data on Composite Asphalt Pavement 190