Dit proefschrift is goedgekeurd door de promotor: Prof. Dr. H. J. van Zuylen Samenstelling promotiecommissie:

Transcription

1

2

3 Huizhao Tu()

4 Dit proefschrift is goedgekeurd door de promotor: Prof. Dr. H. J. van Zuylen Samenstelling promotiecommissie: Rector Magnificus voorzitter Prof. Dr. H. J. van Zuylen Technische Universiteit Delft, promotor Dr. ir. J. W. C. van Lint Technische Universiteit Delft Prof. Dr. ir. S. P. Hoogendoorn Technische Universiteit Delft Prof. Dr. L. J. Sun. Tongji University, Shanghai, P. R. China Prof. Dr. G. C. de Jong University of Leeds, UK Prof. Dr. P. B. Mirchandani University of Arizona, USA Prof. Dr. C. Witteveen Technische Universiteit Delft This thesis is the result of a Ph.D. study carried out from November 2003 to March 2008 at Delft University of Technology, Faculty of Civil Engineering and Geosciences, Department of Transport and Planning. The research was sponsored in the first year by the Sino-Netherlands ITS training center ( programme and from 2005 onwards by the ATMO (Advanced Traffic MOnitoring) research project ( under the Dutch national traffic and transport research program TRANSUMO (TRAnsition SUstainable MObility -- TRAIL Thesis Series no. T2008/7, The Netherlands TRAIL Research School Published and distributed by: TRAIL Research School P. O. Box GA Delft The Netherlands T: +31 (0) F: +31 (0) E: info@rstrail.nl ISBN: Keywords: Travel time reliability, Travel time uncertainty, Instability, Traffic breakdown, Dynamic traffic assignment Cover illustration: t t 2008 by Han Feng Copyrightt t 2008 by Huizhao Tu() All rights reserved. No part of the material protected by this copyright notice may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and retrieval system, without written permission of the author. Printed in The Netherlands

5

6

7

8 Huizhao Tu ()

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25 q on (veh/h) % confidence interval mean q on regression q on q (veh/h) in

26

27

28 Chapter 2 Fundamental Notions of Travel Time Reliability: General Overview Chapter 3 Conceptual Travel Time Reliability Model Chapter 4 Inflow-Travel Time Reliability Model Chapter 5 Extended Travel time Reliability Model Chapter 6 Model Applications

29

30 Point B distance section k-1 section k section k+1 V k+1(p) V k(p) V k-1(p) V k(p+1) V k-1(p+1) V k+1(p+1) Point A t 0 period p Section-level travel time period p+1 time Instantaneous travel time Dynamic (experienced) travel time Route-level

31

32

33

34

35

36

37

38

39 Demand on parallel and / or connecting infrastructure, spillback effects Seasonal effects: time of day, day-of-week, month Incidents and accidents, roadworks Weather, luminance Population characteristics, traffic mix, cultural factors Demand fluctuations Supply fluctuations Road geometry, and regulations Traffic information and user response Traffic operations on Infrastructure network (dynamic) traffic management and control Travel time distribution

40 traffic volume (veh/h) Mean traffic volume on A12 freeway in southeast part of The Netherlands 5500 July 5000 August October :00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 departure time

41 Inflow (veh/hour) route1201(working days, 2004) q10th qmean q90th 0 06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22: route1201(weekend days, 2004) q10th qmean q90th :00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 Departure time (in 10 minutes)

42

43 Capacity (veh/h) 6000 Capac ity on A 15 freeway (veh/h) Tim e s lots ) (

44 Traffic demand Infrastructure supply (Dynamic) traffic assignment model o Equilibrium conditions o (Link) travel time functions - fundamental diagram o... Outputs o Link and route flows o Link travel times o Route travel costs (times) o... Travel time distribution

45

46

47

48

49

50 Traffic data collection Outputs o Traffic flow o Travel speed o Weather conditions o Road geometry o Roadworks o (dynamic) Traffic information o... Data processing Travel time distributions Travel time reliability analysis

51

52 Travel times Travel time reliability Other travel costs Route and departure time choice Flows

53 Characteristics of travel time-flow plane and speed-flow plane (section 3.2) Elements of travel time reliability (section 3.3) Uncertainty in travel times Travel time Variability Instability of travel times Probability of traffic breakdown Travel time reliability model (section 3.4) Travel Time UnReliability model

54

55 v 0 Free flow Speed V Congested flow Flow q q max

56 travel time (seconds/km) speed (km/h) 150 (a)freeway speed-flow relationships λ t λ c (b)freeway travel time-flow relationships λ t λ c traffic volume (veh/h/ln)

57 Uncertainty Instability q in λ t λ c

58

59

60

61 B speeds (km/h) Loop detector location (m) B B B :00 14:00 16:00 18:00 20:00 Departure time(in 10-minute) 20 0

62 q in q on Section one (k 1 ) Section two (k 2 ) Route (r)

63 probability of traffic breakdown Probability of traffic breakdown on A1201 (2004) p br k1 (section one k 1 ) p br k2 (section two k 2 ) P br (estimated from Eq. 3.7) r P br (estimated from Eq. 3.8) r inflow(veh/h/ln)

64

65 travel time (seconds/km) 10 3 Freeway travel time-flow relationships after breakdown before breakdown traffic volume (veh/h/ln)

66 Travel time uncertainty (s/km) Uncertainty after breakdown Uncertainty before breakdown q in (veh/h/ln)

67 distance s TT ( q in ) State one Free flow Free flow Free flow period p-1 period p period p+1 period... time distance t TT ( q in ) State two Free flow Breakdown Congested distance period p-1 period p period p+1 period... j TT ( q in ) time State three Breakdown Congested Congested Congested period p-1 period p period p+1 period... time

68

69

70

71

72

73

74 Average journey speed (km/h) Probability density function Probability density function (10-minute interval) Critical capacity flow Critical transition flow 0.02 λ t λ c Inflow (veh/h/ln) 0

75

76

77

78 Traffic data collection systems Other data sources (weather, road works etc.) Raw data Data cleaning and pre-processing Inductive loops detectors Offline travel time estimator Road geometry Input data Historical database Input data Traffic control measure VMS Travel time reliability (TTR) model TTR=f(flow, road geometry, weather, dynamic speed limit ) Inflow

79

80

81

82

83

84 Percentile travel time (seconds/km) TT10th TT50th TT90th (a) A1201 (2004) Critical capacity inf lowλ c Critical transition inf lowλ t TT10th TT50th TT90th (b) A2001 (2004) Critical capacity inf lowλ c Critical transition inf lowλ t Inf low (v eh/h/ln,10-minute aggregate)

85

86 probability of traffic breakdown 1 (a) A1201 (2004) (b) A1211 (2004) inflow (veh/h/ln, 10-minute aggregate)

87 Travel time variability (seconds/km) TTV f TTV j (a)a1201 (2004) TTV f TTV j (b)a1211 (2004) Inflow (veh/h/ln)

88 Travel time unreliability (seconds/km) A1201 (2004) CTTR TLZ function A1211 (2004) A1301 (2004) A1501 (2004) A2001 (2004) A2011 (2004) Inflow (veh/h/ln)

89

90 Probability of traffic breakdown Speed (km/h) 80 Speed-flow relation (September 1st, 2006) Volume (veh/h) Beijing urban freeway (second ring north part) probability of traffic breakdown inflow (veh/h/ln)

91 Travel time variability (seconds/km) TTV f TTV j Beijing urban freeway (second ring north part) inflow (veh/h/ln)

92 Travel time unreliability (seconds/km) Beijing urban freeway (second ring north part) CTTR TLZ function inflow (veh/h/ln)

93 Travel time unreliability (seconds/km) Probability of traffic breakdown estimated observ ed (a) A12 f reeway (east direction) estimated observ ed (b) A12 f reeway (west direction) inf low (veh/h/ln, 10-minute aggregate) estimated observed (a) A12 freeway (east direction) estimated observed (b) A12 freeway (west direction) Inflow (veh/h/ln, 10-minute aggregate)

94

95 probability of traffic breakdown probability of traffic breakdown under different time intervals 5 - minute 10-minute 15-minute 20-minute 30-minute inflow (veh/h/ln)

96 travel time unreliability (seconds/km) fitted travel time unreliability under different time intervals 5-minute 10-minute 15-minute 20-minute 30-minute inflow (veh/h/ln)

97

98

99

100

101

102 TTV and TTUR versus Freeway characteristics TTV: Travel Time Variability TTUR: Travel Time UnReliabiity TTV and TTUR versus Basic freeway segment TTV and TTUR versus Ramp section TTV and TTUR versus Weaving section Inflow Inflow Inflow Inductive loops detectors Basic freeway segment Ramp section Weaving section

103 (a) Type A (b) Type B (c) Type C

104

105

106 TTUR (s/km) p TTV (s/km) TTUR (s/km) p TTV (s/km) TTV f TTV j probability of traffic breakdown TTUR = x La; = 5.8, La<300m La 300m measured TTUR regression TTUR Length of acceleration lane(m) TTV f TTV j probability of traffic breakdown TTUR = x Ld, Ld<250m measured TTUR 10 =4.4, Ld 250m regression TTUR Length of deceleration lane(m)

107 TTUR (s/km) p TTV (s/km) TTV f TTV j probability of traffic breakdown TTUR = x Lw, =3.6; Lw <750m Lw 750m measured TTUR regression TTUR Weaving length (m)

108 (a) Travel time variability (seconds/km) b1208 b2018 TTV f b2008 TTV j (b) Probability of traffic breakdown b1208 b2018 b (c) Travel time unreliability (seconds/km) 5 b1208 b2018 b Length of basic freeway segments

109 Critical inflow (veh/h/ln) measurements measurements estimated λ t estimated λ c λ t λ c 1800 λ c = -127* β % confidence λ t 95% confidence λ c λ t = -111* β β: number of ramps per ten kilometer Number of ramps per ten kilometers

110 (a) Trav el time v ariability (seconds/km) TTV f TTV j (b) Probability of traf f ic breakdown (c) Trav el time unreliability (seconds/km) Number of ramps per ten kilometers

111

112

113

114 Traffic breakdown probability Traf f ic breakdown probability on A12 f reeway (northern bound) Normal weather Rainy weather Inf low (v eh/h/ln)

115 Travel time (seconds/km) TT10th TT50th Normal Rain TT90th 15 TTUR Weather condition (Rain)

116 Travel time (seconds/km) cumulative function 1 Trav el time distribution (A12 f reeway, east direction) normal weather rain weather trav el time (seconds/km) TT10th TT50th Normal Snow TT90th 15 TTUR Weather condition (Snow)

117

118 Travel time (seconds/km) TT10th TT50th Normal Fog TT90th 15 TTUR Weather condition (Fog)

119 Travel time (seconds/km) TT10th TT50th Normal Ice TT90th 15 TTUR Weather condition (Ice)

120 Travel time variability(seconds/km) Travel time (seconds/km) TT10th TT50th Normal Storm TT90th 15 TTUR Weather condition (Storm) Normal Rain route Normal Rain route Normal Rain route Normal Rain route Normal Rain route Normal Rain route inflow in 10 minutes (veh/h/ln)

121 Travel time unreliability(seconds/km) route1201 Normal Rain route route route route route inf low in 10 minutes (veh/h/ln)

122

123 120km/h area 80km/h area

124

125 Probability of traffic breakdown Speed limit(120km/h) Speed limit(100km/h) A13 f reeway (south direction) Speed limit(120km/h) Speed limit( 80 km/h) A20 f reeway (east direction) inf low (veh/h/ln, 10-minute aggregate)

126 Travel time variability (seconds/km) Travel time variability (seconds/km) Speed limit(120km/h) Speed limit(100km/h) TTV f TTV j Inf low (v eh/h/ln,10-minute aggregate, A13 f reeway (south direction) Speed limit(120km/h) Speed limit( 80 km/h) TTV f TTV j Inf low (v eh/h/ln,10-minute aggregate, A20 f reeway (south direction)

127 Travel time unreliability (seconds/km) Speed limit(120km/h) Speed limit(100km/h) A13 f reeway (south direction) Speed limit(120km/h) Speed limit( 80 km/h) A20 f reeway (east direction) inf low (veh/h/ln, 10-minute aggregate)

128

129

130

131 probability of traffic breakdown Travel time variability (seonds/km) without accidents with accidents TTV f without accidents with accidents TTV j inf low (veh/h/ln) without accident with accident inf low (veh/h/ln)

132 travel time unreliability (seconds/km) without accidents with accidents inf low (veh/h/ln)

133

134

135

136

137

138

139 Link 1 (route one) o d Link 2 (route two)

140 travel time unreliability (minute) departure flow rate (veh/minute) without reliability (route one) without reliability (route two) with reliability (route one) with reliability (route two) departure time (in minute ) 10 9 route one route two departure time (in minute )

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166 Space (x) The trajectory method requires a space-time grid with rectangular regions {k,p} K k {k,p} Each region has up and downsteam detectors producing time-averaged speeds each periodp 1 1 p P Time(t)

167

168 Start vehiclei, set k (usually first section), and p (depends on resolution), and consequently set {x i 0,t i 0 } Vehicleienters region {k,p} at location {x i 0,t i 0 } {x i 0,t i 0 }= {x i *,t i * } Calculate exit-point {x i *,t i * } (with use of section level travelt time estimation!!!) False x i *=x i True k=k+1 k>k False True False t i *=t i True p=p+1 p>p True End of trajectory of vehiclei, record its departure time and path travel time...

169

170

171

172

173

174

175

176

177

178

179 (a) (b) (c) (a)()(b) (c) breakdown

180 CTTR CTTR spontaneous2breakdowninduced2breakdown C T T R C T T R Regiolab-Delft CTTR TLZTu, vanlint, vanzuylen TLZ CTTR

181 8 TLZ TLZ TLZ TLZ Summary (Chinese version) has been approved by Prof. dr. Lijun Sun:

182

183 Huizhao Tu () was born in Changting, Fujian Province, China, on April From 1996 to 2003, he studied at Tongji University where he received his BSc degree ( with honor) in Civil Engineering in 2000 and his MSc degree in Transportation Engineering in 2003 under the supervision of Professor Lijun Sun. During his studies in Tongji he was involved in several projects, among which, Feasibility Study on Implementing Electronic Road Pricing in Shanghai (Shanghai Sci-Tech Progress Award II, 2004), Shanghai Infrastructure Management Systems (Shanghai Sci-Tech Progress Award III, 2003), Ten-Year Transport Strategy Development on Urban Road in Shanghai (Shanghai Sci-Tech Progress Award III, 2005), and Traffic Evaluation of Shanghai Urban Freeways, Traffic Data Collection and Processing for Shanghai. His master thesis concerned the Urban Roads Evaluation Model and its Implementation on Electronic Road Pricing in Shanghai, defended successfully in In 2002 he attended an ITS training course in Shanghai, where he met his current supervisor, Professor Dr. Henk van Zuylen.involved in In November 2003, Huizhao Tu moved to The Netherlands to start his PhD at the Department of Transport and Planning, Delft University of Technology and TRAIL Research School. His PhD research is part of the ATMO project, short for Advanced Traffic MOnitoring. ATMO is one of the projects in the cluster Traffic Management of the TRANSUMO program (TRANsition SUstainable MObility). The central objective of ATMO is the development of knowledge and skills in the structural setup, realization, maintenance and operation of a robust and reliable monitoring system for heterogeneous traffic networks that generates (predicted) traffic information for both traffic managers and traffic participants. This research, conducted under the supervision of Professor Dr. Henk van Zuylen and Dr. Ir. Hans van Lint, resulted in a number of publications in proceedings of several international conferences, for instance TRR, TRB, IEEE ITSC, IFAC, and INSTR. His PhD dissertation addresses Monitoring Travel Time Reliability on Freeways. After he finished his PhD in April 2008, he started working for Grontmij, an international consultancy company in The Netherlands.

184

185

186

187

188

189

190

191

192