Pilot Testing of SHRP 2 Reliability Data and Analytical Products: Florida (2014) / Chapter Skim
Currently Skimming:
Pages 129-147
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 129... ...
129 CHAPTER 4 L07 Product Tests 4.1 Introduction and Background Project L07 focused on the estimation of the effect of physical design treatments on freeway travel time reliability and the cost-effectiveness of these treatments. The L07 analysis was based on the reliability estimation methods developed in the L03 project and can be considered as a sketch planning–level analysis that is not as detailed as the L08 procedures.
|
From page 130... ...
130 Table 4.1. Coefficients Used in Project L03 Reliability Models for Peak Hour N (percentile)
|
From page 131... ...
131 LHL due to shoulder blockage can also contribute to incident impacts. This contribution was accounted for by calculating the equivalent lane blockage due to reduced capacity as a result of shoulder blockage multiplied by the blockage duration.
|
From page 132... ...
132 procedure was applied in this study by using a mean speed of 40 mph as a threshold to activate the congested demand–estimation procedure. 4.3.1.3 Hours of Rainfall Exceeding 0.05 inch The L07 tool provides default values of hourly rainfall based on 10 years (2000 through 2010)
|
From page 133... ...
133 Figure 4.2. Cumulative TTIs for Segment 1 of I-95 NB GPL.
|
From page 134... ...
134 Figure 4.4. Cumulative TTIs for Segment 3 of I-95 NB GPL.
|
From page 135... ...
135 Figure 4.6. Comparison of mean, 95th percentile, and 80th percentile TTIs for Segment 1 of I-95 NB GPL.
|
From page 136... ...
136 Figure 4.8. Comparison of mean, 95th percentile, and 80th percentile TTIs for Segment 3 of I-95 NB GPL.
|
From page 137... ...
137 4.4 Parameter Estimation Based on Local Conditions As mentioned in Section 4.3, the default TTI model used in L07 did not produce a good estimation of the TTIs for the I-95 NB GPL. This section presents the derivation of the parameters of the TTI estimation model based on local data.
|
From page 138... ...
138 Figure 4.10. Scatterplots for precipitation and segment length.
|
From page 139... ...
139 Table 4.4. Comparison between Global Regression Models Fit of Real-World Data Model R2 Format RMSE CV (%)
|
From page 140... ...
140 Table 4.5. Comparison between Different Individual Regression Models Model R2 Format RMSE CV Segment 1 0.708 0.192441 14.7% 0.990 0.035533 2.7% Segment 2 0.725 0.134105 10.9% 0.970 0.442901 3.6% Segment 3 0.665 0.087111 7.8% 0.977 0.022445 2.0% Table 4.6.
|
From page 141... ...
141 Figure 4.11. Comparison of regression models for Segment 1.
|
From page 142... ...
142 Figure 4.13. Comparison of regression models for Segment 3.
|
From page 143... ...
143 negative values. The updated regression analysis for the mean TTI is described below as an example.
|
From page 144... ...
144 Figure 4.15. Comparison of regression models for Segment 2.
|
From page 145... ...
145 Similar analysis was also conducted for other TTI percentiles in addition to mean TTI. Tables 4.9 and 4.10 show the regression coefficients and R 2 values for different models without constraining the coefficient signs to positive values and after constraining the signs to positive values, respectively.
|
From page 146... ...
146 Individual model for Segment 3: Percentile R 2 b1 b2 b3 b4 b5 10 0.931 –5.148 0.253 –0.111 –5.192 1.006 50 0.980 –20.766 0.508 –0.275 4.153 1.032 80 0.983 –9.543 0.387 –0.287 –1.873 1.049 95 0.953 52.466 –0.243 –0.192 –41.800 1.074 99 0.893 74.582 –0.540 –0.044 –54.379 1.114 Mean 0.977 4.655 0.185 –0.197 –10.143 1.026 Table 4.10. Coefficients for Different Models after Validation Global Model 4: Percentile R 2 b1 b2 b3 b4 b5 b6 10 0.581 0.500 0.000 0.013 –0.075 –1.555 0.749 50 0.864 17.445 0.000 0.000 –2.457 –15.568 1.071 80 0.825 14.865 0.000 0.000 –0.658 –13.912 1.072 95 0.827 10.477 0.029 0.000 –0.832 –9.139 1.105 99 0.814 5.481 0.049 0.000 –0.894 –3.758 1.105 Mean 0.884 14.020 0.000 0.000 –0.619 –13.470 1.058 Global Model 5: Percentile R 2 b1 b2 b3 b4 b5 b6 b7 10 0.564 0.271 0.000 0.009 –0.954 0.029 –0.011 0.536 50 0.685 13.978 0.000 0.000 –14.684 –0.447 0.078 1.514 80 0.698 12.802 0.000 0.000 –12.987 –0.782 0.139 1.880 95 0.733 10.757 0.000 0.000 –10.426 –0.626 0.087 1.864 99 0.757 6.791 0.000 0.000 –5.864 –0.594 0.046 1.986 Mean 0.757 12.101 0.000 0.000 –12.632 –0.410 0.066 1.509 Individual model for Segment 1: Percentile R 2 b1 b2 b3 b4 b5 10 0.791 0.000 0.246 0.015 –3.180 0.930 50 0.921 16.961 0.517 0.000 –20.256 1.095 80 0.857 0.000 1.641 0.000 –9.751 1.135 95 0.821 0.000 1.123 0.000 –6.167 1.133 99 0.841 0.000 0.579 0.000 –2.328 0.922 Mean 0.910 0.000 1.584 0.000 –9.786 1.101
|
From page 147... ...
147 Individual model for Segment 2: Percentile R 2 b1 b2 b3 b4 b5 10 0.625 0.000 0.044 0.000 –1.727 0.817 50 0.910 8.326 0.318 0.000 –13.488 1.047 80 0.890 11.483 0.122 0.000 –13.318 1.047 95 0.885 9.529 0.097 0.000 –10.451 1.134 99 0.833 0.000 0.175 0.000 –1.276 0.789 Mean 0.925 10.528 0.135 0.000 –12.885 1.062 Individual model for Segment 3: Percentile R 2 b1 b2 b3 b4 b5 10 0.639 0.000 0.006 0.000 –0.949 0.589 50 0.789 0.000 0.228 0.000 –10.039 1.034 80 0.751 0.000 0.222 0.000 –9.081 1.049 95 0.793 16.848 0.000 0.000 –17.221 1.066 99 0.761 11.081 0.000 0.000 –10.675 1.106 Mean 0.823 9.018 0.087 0.000 –13.579 1.024 4.5 Summary The L07 tool was applied in this study to investigate the travel time reliability along the I-95 NB segments. The analysis results showed that the TTI prediction model used in the L07 tool was more sensitive to the incident's number and duration than other variables, such as traffic demand or weather.
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.