37
Draft Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Journal: Canadian Journal of Civil Engineering Manuscript ID cjce-2018-0601.R1 Manuscript Type: Article Date Submitted by the Author: 15-May-2019 Complete List of Authors: Lai, Yuanwen; Fuzhou University, College of Civil Engineering Xu, Xinying; Fuzhou University, College of Civil Engineering Easa, Said; Ryerson University, Department of Civil Engineering Lian, Peikun; Fujian Agriculture and Forestry University, College of Transportation and Civil Engineering Keyword: Bus signal priority, modeling, signal timings, mobile-phone GPS data, micro simulation Is the invited manuscript for consideration in a Special Issue? : Not applicable (regular submission) https://mc06.manuscriptcentral.com/cjce-pubs Canadian Journal of Civil Engineering

Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

  • Upload
    others

  • View
    5

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data

Journal: Canadian Journal of Civil Engineering

Manuscript ID cjce-2018-0601.R1

Manuscript Type: Article

Date Submitted by the Author: 15-May-2019

Complete List of Authors: Lai, Yuanwen; Fuzhou University, College of Civil EngineeringXu, Xinying; Fuzhou University, College of Civil EngineeringEasa, Said; Ryerson University, Department of Civil EngineeringLian, Peikun; Fujian Agriculture and Forestry University, College of Transportation and Civil Engineering

Keyword: Bus signal priority, modeling, signal timings, mobile-phone GPS data, micro simulation

Is the invited manuscript for consideration in a Special

Issue? :Not applicable (regular submission)

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 2: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

1

Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data

Yuanwen Lai,1 Xinying Xu,2 Said M. Easa,3 and Peikun Lian4

1 Yuanwen Lai, College of Civil Engineering, Fuzhou University, Fuzhou, China

2 Xinying Xu, College of Civil Engineering, Fuzhou University, Fuzhou, China.

3 Said M. Easa, Department of Civil Engineering, Ryerson University, Toronto, Canada.

4 Peikun Lian, College of Transportation and Civil Engineering, Fujian Agriculture and

Forestry University, Fuzhou, China.

Corresponding Author: Xinying Xu

College of Civil Engineering, Fuzhou University, Fuzhou, China,

Tel: +8615605082423, E-mail: [email protected].

The manuscript consists of 5706 words.

Page 1 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 3: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

2

Abstract

Limited by the low-frequency data acquisition, vehicle global positioning system (GPS) data

are difficult to implement in the area of micro-traffic simulation. Based on the functional

design of mobile-phone positioning technology, mobile phones can be used to acquire bus

GPS data every second. In this paper, an analytical model is proposed to determine the

parameters of signal coordination for bus priority along an arterial based on GPS data of

mobile phones. First, bus priority evaluation indicators are established using bus GPS data

which are acquired by mobile phones. Second, the signal timing parameters of the arterial

road are optimized, and a preliminary timing plan is developed by evaluating small changes in

the plan. Finally, the corresponding final plan is developed using VISSIM micro simulation

software. The feasibility of the analytical model is verified by simulating an actual arterial in

Fuzhou city, China.

Key Words: Bus signal priority; modeling; signal timings; mobile-phone GPS data; micro

simulation.

Page 2 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 4: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

3

1 1. Introduction

2 With the rapid development of urbanization and increasing demand on automobile, traffic

3 congestion increasingly grows in most cities. Public transport has the advantages of less

4 energy consumption, large capacity, and low cost. The implementation of public transport

5 priority mode has become an effective means to alleviate traffic congestion. Bus priority

6 control strategies can be divided into passive priority, active priority, and real-time priority,

7 according to Byrne et al. (2005). Passive priority strategy only optimizes the off-line timing

8 program and does not need to detect whether there is a bus arrival. Active priority strategy

9 checks whether there is a bus arrival and determines whether to give a priority signal. In

10 real-time priority strategy, the bus priority signal is given based on real-time detection data,

11 and an objective function is developed to optimize a signal timing plan. In recent years,

12 researchers have focused attention on active priority and real-time priority strategy, while

13 active priority strategy is more widely used in real-life traffic signal control.

14 For isolated intersection control, Vincent et al. (1978) used micro simulation to propose

15 five active priority strategies: (a) only green interval is extended, (b) green interval is

16 extended and red interval is shortened, (c) both green and red intervals are extended along

17 with a recovery algorithm (an algorithm that returns to the original signal timing), (d) red

18 interval is shortened, and (e) red interval is shortened along with a recovery algorithm. Xu et

19 al. (2008) developed a control strategy based on the rules for priority generation request, rules

20 for green time adjustment, and rules for barrier crossing. Yu et al. (2015) developed a fuzzy

21 rule for the bus priority active control model. Given traffic conditions, phase release order,

22 and a green interval three-level fuzzy controller, the model determined two priority strategies

23 to advance phase release order and extend green interval. Wang et al. (2016) developed a bus

24 priority control strategy and evaluation method based on the overall delay at intersections.

25 Tien et al. (2018) presented an advanced transit signal priority (ATSP) control model that

26 considered bus progression at downstream intersections when giving priority at upstream

27 intersections along with stochastic bus arrival times. In general, the basic methods of active

28 priority strategy have been addressed from the perspective of rules and models.

29 Arterial signal coordination for bus priority has been investigated by several researchers.

Page 3 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 5: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

4

30 Skabardonis (2000) used passive priority and active priority for arterial bus coordination

31 control. Balke et al. (2000) proposed an arterial bus priority control algorithm that did not

32 affect public transit vehicles or signal timings. The algorithm included four steps: arrival time

33 prediction, priority judgment, strategy choice, and strategy implementation. Meenakshy (2005)

34 presented a bus priority timing model that considered green wave coordination control as the

35 premise and the average delay as an evaluation indictor. Li et al. (2015) developed an active

36 priority strategy for bus priority along an arterial that minimized the total weighted delay of

37 adjacent intersections. Zhen Y (2015) developed an intersection signal timing model based on

38 genetic algorithms that included start-up wave transmission line in the traditional delay

39 triangulation method. The model can be used to determine the duration of cycle length and

40 effective green time in bus coordination control as well as phase difference and phase

41 sequence, considering the effects on bus trip time and number of stops. Khaled and

42 Mohammad (2018) evaluated the potential benefits of implementing transit signal priority

43 along a major corridor in the City of Doha using VISSIM multimodal microsimulation

44 (abbreviations refer to German words, meaning Traffic in cities - simulation model). The

45 results show that travel time was reduced up to 43% in some cases and this can be translated

46 into lower transit delay and more reliable transit service. As can be seen, there is much

47 research work on parameter optimization of bus priority control, but most optimization

48 methods are applicable to specific traffic conditions. Real-life traffic conditions, however,

49 constantly change. Therefore, it is necessary to develop a model to improve the adaptability

50 of the timing plans for the changing traffic conditions.

51 At present, the design of bus timing control strategy mainly depends on some detector

52 data, such as radar data and induction coil data. Therefore, it is difficult to adapt discrete

53 changes in traffic flow to such data. Using GPS technology, the data can be constantly

54 updated. Many researchers have focused their studies on bus GPS data. Hounsell et al. (2007)

55 combined the door-closing sensor and virtual detector to tackle the challenge posed by the

56 locational error associated with GPS where a traffic signal is close to a bus stop. Eirikis et al.

57 (2010) used GPS data to propose a method that can capture buses’ real-time location to

58 reduce passengers’ waiting time at the site. Wei et al. (2015) used GPS data to develop a

Page 4 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 6: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

5

59 dynamic travel time prediction model. Wang et al. (2014) developed an algorithm to calculate

60 the acceleration and deceleration of bus at intersections based on low-frequency GPS data.

61 Wang et al. (2015) developed a reliability evaluation system of bus travel time using a bus

62 GPS data map matching. Gao (2017) proposed a key technology for bus priority signal

63 control at intersections based on bus GPS data. Currently, the sampling interval of vehicle

64 GPS data is generally 10 s to 45 s. Limited by the low frequency of acquisition, vehicle GPS

65 data is mainly used for such applications as traffic conditions forecast, bus route planning,

66 and bus operation scheduling, while the data are less used for intersection signal optimization

67 and evaluation.

68 With the design and development of mobile phone positioning technology, through

69 programing and smart phone applications, GPS data of mobile phones can avoid the problem

70 of low sampling frequency, so that feedback adjustment becomes possible with the GPS data

71 applied to the signal control mode. Many studies about mobile phone GPS data have been

72 recently conducted, such as application to predicting travel time (Woodard et al., 2017) and

73 identifying travelers' transportation modes (Zhou et al., 2018). By means of functional design

74 and development of mobile phone GPS, the Beijing traffic management department has

75 recently reduced GPS data collection frequency to 1 s and used the data to evaluate the effect

76 of implementing arterial coordination.

77 This paper proposes a new method for optimizing the parameters of arterial bus

78 priority control using mobile-phone GPS data. The contributions of this paper are as follows:

79 (a) introducing mobile-phone GPS implementation that improves the frequency of GPS data

80 acquisition;(b) establishing evaluation indicators of arterial bus priority using mobile-phone

81 GPS data;(c) developing an analytical model for determining the plan of arterial bus priority

82 that best adapts to dynamic traffic flows; and (d) verifying the proposed model using a

83 simulation experiment of an actual arterial.

84 The remainder of the paper is organized as follows. Section 2 presents the development

85 of the model, including defining evaluation indicators, determining initial arterial signal

86 parameters, generating initial timing plans, and selecting matching plan. Section 3 presents an

87 application of the model using simulation of an actual arterial. Section 4 presents the

Page 5 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 7: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

6

88 conclusions.

89 2. Model Development

90 The development of the model involves the following assumptions:

91 (a) The model considers a series of intersections along the arterial that meet the conditions for

92 setting the arterial coordination control.

93 (b) Only the peak hours of a typical working day are considered.

94 (c) The intersections of the arterial adopt a single-cycle balance strategy,such as “signal red

95 light early break-off”. In this strategy, when it comes to bus priority, the green time of the

96 priority phase is compensated by a reduction of the non-preferential phase to ensure that

97 the cycle length remains the same.

98 (d) The model only considers the situation where the bus arrives at the green wave

99 coordination phase, and the bus phase is the arterial coordination phase.

100 The design concept of this study is based on acquiring bus GPS data using mobile

101 phones and obtaining the priority evaluation indicators of the bus. Then, based on the

102 single-cycle balance strategy of the Hisense SC3080 controller, signal timing parameters are

103 optimized. The process of parameter optimization of arterial bus priority coordination control

104 is shown in Fig. 1. The analytical model for determining the best plan of bus priority involves

105 three tasks: (1) preliminary parameters of arterial coordination control are determined, (2)

106 multiple plans are designed based on the preliminary parameters, and (3) the best timing plan

107 is determined according to the adaptability of the parameters to changes in traffic flow.

108 2.1 Defining Evaluation Indicators

109 The evaluation indicators of arterial bus priority include travel time, number of stops, and bus

110 delay. These indicators, which reflect the quality of coordination of traffic signals along the

111 arterial and the efficiency of bus operation, can be easily determined from mobile-phone GPS

112 data. Before describing the evaluation indicators, it is useful to describe the mobile-phone

113 data acquisition method.

114 2.1.1 Mobile-Phone Data Acquisition

115 The acquisition method is mainly performed manually to obtain the location information of

116 the bus on the arterial road. The investigator carries a mobile phone to take the designated bus,

Page 6 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 8: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

7

117 and an app that can acquire GPS data is installed on the mobile phone. Through the location

118 function of the mobile app, the latitude and longitude coordinates of the investigator are

119 automatically obtained every second. The coordinates of the investigator are regarded as the

120 real-time location of the bus.

121 The VISSIM software uses the world coordinate system by default. Each simulation

122 point on the system has fixed XY coordinate values. Therefore, the VISSIM simulation

123 coordinate values were used to simulate the bus latitude and longitude values. The software is

124 written using the COM interface (a technology that enables inter-process communication

125 between software), and the evaluation indicators of arterial bus priority extracted from the

126 simulation coordinate data were imported into the Access database as a test data support for

127 further processing and analysis. The COM interface was used to write related programs to

128 extract bus-priority evaluation indicators from the simulation coordinate data, then importing

129 the simulation data into the Access database as the test data.

130 2.1.2 Bus Travel Time

131 A 1-hour test duration is used in this study, where the traffic condition is expressed as the

132 average value of the traffic parameters during the hour. Let k denote the number of tests. In

133 test k, there are nk buses equipped with GPS receivers that run through the entire arterial.

134 Then, the average travel time of all buses in test k is given by

135 (1)

k

n

jk

k n

jtt

k

1

)(

136 (2))()()( jtjtjt kakdk

137 where is average travel time of all buses in test k, where k = 1,2, …, m; is travel kt ( )kt j

138 time of bus j in test k, where j = 1, 2, ..., nk; )( jtka is time when bus j enters the arterial,

139 where j = 1, 2, ..., nk; and )( jtkd is time when bus j exits the arterial, where j = 1, 2, ..., nk.

140 2.1.3 Number of Stops and Bus Delay

141 After a simple processing of the bus GPS data, the number of bus stops is determined. When

142 the speed of the bus in the GPS data is zero, the vehicle is in a stop state. During a period, the

Page 7 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 9: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

8

143 number of stops in a test can be recorded and all times of the bus stops in a test is added up as

144 bus delay. The average number of bus stops on the arterial is calculated as the total number of

145 stops for all buses in test k divided by the number of buses. That is,

146 (3)

k

n

jk

k n

jNn

k

1

)(

147 where is average number of stops for all buses in test k, j = 1, 2, ..., nk and is kn )(k jN

148 number of stops of bus j in test k.

149 Similarly, the average stop delay on the arterial is the average of the delay for all buses

150 in test k, which is given by

151 (4)

k

n

jk

k n

jtt

k

1

)(

152 where is average delay for all buses in test k, where j = 1, 2, ..., nk and is kt )( jtk

153 delay of bus j in test k.

154 The displacement time series plot of bus j in test k is graphically shown in Fig. 2. In this

155 figure, the travel time of bus j in test k, , is the difference between and . ( )kt j ( )dt j ( )at j

156 The number of stops of bus j in test k, , is the number of horizontal bold lines in the )(k jN

157 diagram. The delay of bus j in test k, , is the total length of the horizontal bold lines. )( jtk

158 2.1.4 Deviation of Bus Flow

159 To ensure a two-way bus priority, the service quality of the upstream and downstream bus

160 flows is balanced using a new overall measure called the deviation of bus flow, which is

161 defined as follows

162 (5)

3

))() 222

kt

tkn

nkt

tS

kk

kk

kk

((

163 where is degree of deviation between the upstream and downstream bus indicators and the S

164 average values of the indicators.

Page 8 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 10: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

9

165 2.2 Determining Initial Arterial Signal Parameters

166 2.2.1 Cycle Length

167 The optimal cycle length of each intersection on the arterial is calculated according to the

168 following Webster’s formula,

169 (6)

1.5 51

ii

i

LcY

170 (7)icC max

171 where is cycle length of intersection (s), Li is total green lost time of intersection (s); ic i i

172 Yi is total traffic flow ratio of intersection (sum of critical lane flow to saturation flow i

173 ratios); and is common cycle for traffic signal coordination along the arterial (s).C

174 2.2.2 Green-Phase Ratio

175 The effective green time of phase j at intersection i is calculated according the ratio of critical

176 flow to saturation flow. That is,

177 (8))( i

i

jiji LC

Yy

g

178 (9)C

g jiji

179 where gji is effective green time of phase j at intersection i (s); yji is critical flow to saturation

180 flow ratio of phase j at intersection I; and λji is green-phase ratio of phase j at intersection i;.

181 2.2.3 Phase Difference

182 When bus priority is not considered (i.e. green-phase ratio is fixed), several methods can be

183 used to determine the two-way green wave bandwidth and phase difference, such as graphical

184 method, mathematical method, genetic algorithm, or MAXBAND method (a computer

185 software for setting arterial signals to achieve maximal bandwidth) can be used (Gartner 1991).

186 For the case of bus priority, unequal bandwidth can be determined using optimization based

187 on actual intersection spacing, average speed, and actual traffic flow at the intersections.

188 2.3 Single-Cycle Bus Priority Control Logic

189 The single-cycle control logic is suitable for bus priority to arterial signal coordination and

Page 9 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 11: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

10

190 mainly adopts green extension and early green strategies. For the priority phase, the red light

191 is prematurely broken, or the green light is turned on early, while for the non-priority phase,

192 the green light is prematurely broken. The non-priority phase premature break time equals the

193 priority phase extension time, so that the cycle duration does not change. For the non-priority

194 phase premature break module, the premature break time is the time when the detector detects

195 bus application. The logic of this strategy is shown in Fig. 3.

196 2.4 Generating Initial Timing Plans

197 After the initial optimization of signal timing parameters, initial timing plans are generated by

198 introducing small proportional variations in the values of the preliminary parameters, as

199 follows:

200 1. Common Cycle Plan: Using the common cycle length based on Eq. (7), several different

201 common cycles are obtained by adding or subtracting a fixed increment. The increment is

202 selected within 5-10 s to obtain a suitable value. However, the increment should not

203 exceed 10% of the common cycle to maintain a continuous control of the intersections.

204 2. Green-Phase Ratio Plan: The green-phase ratio plan is mainly designed for different traffic

205 flow loads at intersections. Similar to the common cycle plan, several sets of green-phase

206 ratio plans are generated by increasing or decreasing the original green-phase ratio by a

207 small increment. When the priority phase increases the green time, the non-priority phase

208 reduces the green time to ensure that the sum of the green-phase ratios is 1. The increment

209 is selected as 0.05 or less.

210 3. Combined Plan: The preceding common cycle plans and green-phase ratio plans are

211 combined to produce multiple joint plans of common cycle and green-phase ratio.

212 2.5 Selecting Matching Plans

213 The selection of the matching plans involves two tasks: design of traffic scenes and

214 evaluation of timing plans.

215 1. Design of traffic scenes: The traffic system changes with time, so traffic condition at

216 different times exhibits great randomness. It is unreasonable to evaluate the pros and

217 cons of the timing scheme and its timing parameters only based on the traffic data of

218 one survey. In view of this, this paper considers the complexity of traffic and

Page 10 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 12: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

11

219 proposes a method for optimizing the timing parameters under different traffic

220 conditions. The traffic scenes are designed based on the critical influencing factors of

221 traffic conditions and are input for VISSIM to simulate actual traffic conditions.

222 2. Evaluation of timing plans: The evaluation of the timing plans is based on bus priority

223 evaluation indicators. Five ratings are established: Poor, Moderate, Good, Very Good,

224 and Excellent, with ratings of 1 to 5, respectively. For each plan, the rating of each

225 evaluation indicator is determined and the scores of each indicator are added as the

226 score of that plan. The procedures of determining the preferred bus priority timing

227 plan is shown in Fig. 4.

228 3. Application

229 The proposed bus priority model was applied to an actual arterial using the VAP (Vehicle

230 Actuated Programming) module in VISSIM and the Microsoft Visual Studio programming

231 language.

232 3.1 Simulation Environment

233 The study arterial is Jinshan Avenue in Fuzhou city, China. The geometric characteristics of

234 the arterial are shown in Fig. 5. The arterial is a two-way road and consists of six consecutive

235 intersections. The distance between the intersections (stop lines) ranges from 340 m to 800 m.

236 The peak traffic flows and signal timings were the basic simulation parameters.

237 The mainline intersects six north-south roads and passes through six bus stops. The bus

238 routes that cover these six bus stops are 41, 96, 123, and 173. In order to ensure the priority of

239 public transportation, Jinshan Avenue also has a marking bus lane. The bus enters the bus

240 lane in the morning peak (7:00 to 9:00) and the afternoon peak (17:00 to 19:00) to improve

241 the speed of buses. At the end of 2015, Jinshan Avenue has implemented coordinated green

242 wave control, with a green wave speed of 50~55 km/h. The existing signal timing is shown in

243 Fig. 6.

Page 11 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 13: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

12

244 During the afternoon peak hour(17:00-19:00), the actual traffic volume survey was conducted

245 at each intersection and each direction by manual counting. The investigation data were

246 sorted and converted into the equivalent number of passenger cars. The equivalent conversion

247 coefficient was 1 for cars, 1.5 for medium-sized vehicles, and 2 for the large-scale vehicles.

248 The traffic composition was set in the ratio of 8:1:1 for the three models in the VISSIM

249 simulation. The traffic volume observation data after conversion are shown in the Fig. 7.

250 3.2 Plan Design

251 As previously mentioned, the proposed model considers three plans. For the common cycle

252 length, the arterial initial signal timing parameters were used as the basis. Since the base

253 common cycle length was 144 s, five common cycle lengths were considered: 120 s, 130 s,

254 140 s, 150 s, and 160 s. Due to the large pedestrian flow of Jinshan Avenue, in the original

255 timing plan, each intersection has a 20-s pedestrian dedicated green time. Considering bus

256 priority, the 20-s could be taken as the bus priority adjustment phase, so that the common

257 cycle length is adjusted to 140 s, 150 s, 160 s, 170 s, and 180 s. For the plan related to traffic

258 flow levels, four traffic flow levels (-20%, -10%, 10%, and 20%) of the base traffic flow were

259 tested to simulate the effect of the changes in traffic flow. The joint plan consisted of

260 combinations of the common cycle length and traffic flow level.

261 A single-cycle bus priority control logic was coded in VISVAP (Vehicle Actuated

262 Programming of VISSIM). In this study, the evaluation indicators are not directly output by

263 VISSIM simulation system. Instead, the simulation coordinates are used to simulate the

264 mobile-phone GPS data, and then the value is calculated according to the indicator formula.

265 That is, the relevant language is written in Visual Studio to call for the VISSIM_COM

266 module, and the two-way bus simulation data on the arterial were generated second by second.

267 The data were then placed into the bus GPS database, where travel time, delays, and stops of

268 each bus were extracted. A sample of the bus database from VISSIM simulation is shown in

269 Table 1. Table 2 shows the descriptions of the fields of VISSIM simulation bus database

270 shown in Table 1.

271 3.3 Simulation Results

272 3.3.1 Model Effectiveness Analysis

Page 12 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 14: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

13

273 Using VISSIM, the actual traffic data of Jinshan Avenue is used as input data, and the

274 average travel time of all buses, average number of stops of all buses, and average delay of all

275 buses are used as evaluation indicators. The comparative analysis between the single-cycle

276 bus priority control strategy and the arterial green wave coordinated control strategy (no bus

277 priority) is shown in Table 3. As noted, compared to no bus priority, in the single-cycle

278 control strategy the average travel time of all buses is reduced by 2%, the average number of

279 stops of all buses is reduced by 6%, and the average delay of all buses is reduced by 5%.

280 Therefore, this strategy can improve the traffic efficiency of buses.

281 3.3.2 Analysis of Simulation Results of Different Timing Schemes

282 Taking actual traffic data of Jinshan Avenue as an example, the average travel time, average

283 number of stops, and average delay for the five common cycle lengths, obtained from

284 VISSIM simulation, are shown in Fig. 8. Based on the values of the three evaluation

285 indicators, the corresponding scores for the five common cycle lengths were calculated as

286 shown in Table 4.

287 Average Travel Time Analysis

288 As shown in Fig. 8(a), the average travel times for all buses in the five scenarios are

289 compared from west to east (WE), from east to west (EW), and in both directions

290 (TWO-WAY). As noted, for the average travel time of TWO-WAY buses, the average travel

291 time under the timing plan with a common cycle of 180 s is shorter and the timing plan with

292 common cycles of 150 s, and 170 s takes the second place. For the timing plan with common

293 cycles of 140 s and 180 s, the average bus travel time is longer. Obviously, as the common

294 cycle increases, there must be a common cycle that makes the average travel time of the bus

295 the shortest.

296 As for the average travel time of the one-way bus, as the common cycle increases, the

297 average travel time of bus of WE increases, while the average travel time of the bus of EW

298 slightly decreases .

299 Average Number of Stops Analysis

300 As can be seen from Fig. 8(b), the average number of bus stops on the mainline is more than

301 11, with an average of two stops at every two adjacent intersections and on the road between

Page 13 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 15: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

14

302 them. The number of bus stops of WE is less than the number of bus stops of EW. In general,

303 common cycles of 150 s and 160 s have fewer bus stops and are superior to other schemes.

304 Average Delay Analysis

305 The change trend of public parking average delay along with the common cycle and bus

306 travel time are almost the same with the common cycle. Similarly, as the common cycle

307 increases, there must be a common cycle that causes the average public parking delay to be

308 shortest. Bus parking delay is about 520 s, combined with the average number of parking, the

309 average time required for each parking is 47 s.

310 Based on the preceding analysis, a common cycle length of 160 s has the largest score

311 (15) and therefore is the best plan. In addition, from Fig. 8, the values of the indictors for the

312 WE direction of the arterial are smaller than those of the opposite direction, therefore the WE

313 direction is more efficient.

314 3.3.3 Analysis of Simulation Results of Different Traffic Volumes

315 According to the method of timing parameter optimization previously introduced, the effect

316 of the five timing schemes for the five different traffic states is compared and analyzed. The

317 optimal timing scheme is optimized by scoring the pros and cons of the implementation

318 effects.

319 Analysis of Bus Priority Evaluation Indicators

320 The values of the evaluation indicators for the five traffic flows are shown in Fig. 9. For the

321 travel time analysis, it is noted from Fig. 9(a) that when the traffic flow of the mainline

322 increases, the average travel time of buses also increases and at a faster rate. In the case of

323 small flow, the average travel time of the buses under the common cycle 160 s is smaller than

324 that of other timing plans. That is, when the traffic flow is large, the average travel time of the

325 buses under the common cycle of 140 s is higher than that of other timing plans.

326 For parking delay analysis (Fig. 9b), it is noted that the average delay of bus stops

327 increases with the increase of traffic flow and increases more and more. When the traffic flow

328 is small, the average bus stop delay under the timing plan with common cycles of 160 s and

329 180 s is smaller than other timing plans; when the traffic flow at the trunk is large, the average

330 bus stop delay under the 160 s timing plan. Therefore, from the parking delay point of view, a

Page 14 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 16: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

15

331 common cycle of 160 s is the best. For parking frequency analysis (Fig. 9c), the average

332 number of bus stops increases with the increase of traffic flow. Overall, the average number

333 of bus stops under the timing plan with a common cycle of 160 s is less than that of other

334 timing plans. Therefore, based on the analysis of the number of parking trips, the timing plan

335 with a common cycle of 160 s is the best.

336 The evaluation indicators for the plans combining different common cycle lengths and

337 different traffic flows are shown in Table 5. As noted, the common cycle length of 160 s has

338 the largest score (65), indicating that it is more adaptable to changes in traffic flow conditions.

339 Based on the overall analysis, the three indicators of different plans under different traffic

340 flow levels are accumulated, and the results of the final score are shown in Table 4.

341 Comparing the bi-directional three indicators obtained through different timing plans for

342 different traffic flows, it can be concluded that the timing plans with a common cycle of 160 s

343 is the best.

344 Flow Balance Analysis

345 As previously noted, the deviation of the bus flow with a common cycle length of 160 s is

346 smaller than that of the other plans. Therefore, this plan is better as it ensures a two-way

347 balance of traffic flows on the arterial. The deviation of the bus flow is shown in Fig. 10. The

348 analysis of the Jinshan Avenue involved three tasks. First, according to the current traffic data,

349 the common cycle, green-phase ratio, and phase difference were first calculated, then a

350 common cycle is 144 s was obtained, followed by the single-cycle balance plan “signal red

351 light early break off”. Second, according to the calculation results, five timing plans (140 s,

352 150 s, 160 s, 170 s, and 180 s) were designed. It is observed that when the cycle length is 160

353 s, the best results are obtained for arterial bus priority. Finally, by changing the traffic flow

354 level of the arterial, the adaptability of each plan to dynamic traffic is analyzed.

355 4. Concluding Remarks

356 This paper has presented a model that considered bus priority along an arterial with

357 coordinated traffic signals based on GPS data collected from mobile phones. Multiple

358 common cycle lengths and traffic flows were considered using small increments. According

359 to the scores of bus priority evaluation indicators, the model determines the common cycle

Page 15 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 17: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

16

360 length which has the highest ability to adapt to dynamic traffic flow conditions. The specific

361 steps of optimizing arterial bus-priority coordination control based on mobile-phone GPS data.

362 These steps can provide a useful decision-making reference for future studies on urban

363 arterial bus signal priority. Based on this study, the following comments are offered:

364 1. In this study, bus GPS data are collected every second using a mobile phone programming

365 software. Compared with bus GPS data, the acquisition frequency of the mobile software

366 is higher, which can provide stronger data support for the research on micro-traffic

367 simulation. In addition, since the latitude and longitude coordinates of the GPS data and

368 the world XY coordinates system of the VISSIM microscopic simulation software are

369 similar, this software’s coordinate system was used to simulate the latitude and longitude

370 coordinates.

371 2. In evaluating the effectiveness of implementing arterial bus priority, the average bus travel

372 time, number of stops, and bus delay were used as the priority evaluation indicators of

373 arterial buses. In assessing the adaptability of timing plans, the concept of deviation of

374 bus flow is introduced to ensure that the balance and stability of the two-way traffic flow

375 are achieved.

376 3. Taking the Jinshan Avenue arterial coordination as an example, the VISSIM simulation

377 software was used to compare single-cycle bus priority control strategy and arterial green

378 wave coordinated control strategy. The results show that the single-cycle bus priority

379 control strategy is more effectiveness and can improve the traffic efficiency of buses.

380 4. To reduce the complexity of the research problem and highlight the key issues, the

381 proposed model has focused on a single arterial and does not consider the impact of the

382 strategy on social vehicles (vehicles other than buses) or branch buses. The model can be

383 extended to consider the overall benefits of all vehicles. In addition, the number of timing

384 plans can be generated by introducing small proportional variations of the values of the

385 preliminary parameters.

386 Acknowledgements

387 This research is financially supported by the Science and Technology Fund of Education

Page 16 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 18: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

17

388 Department of Fujian Province (JAT160079). The authors are grateful to two anonymous

389 reviewers for their thorough and most helpful comments.

390 References

391 Balke, K.N., Dudek, C.L., and Urbanik, T. 2000. Development and evaluation of an

392 intelligent bus priority concept. Transportation Research Record, 1727, Journal of

393 Transportation Research Board, Washington, D.C., 12-19.

394 Byrne, N., Koonce, P., Bertini, R.L., Pangilinan, C., and Lasky, M. 2005. Using

395 hardware-in-the-loop simulation to evaluate signal control strategies for transit signal

396 priority. Transportation Research Record 1925, Journal of Transportation Research Board,

397 Washington, D.C. 227-234.

398 Eirikis, D. and Eirikis, M. 2010. Friending transit. Mass Transit.

399 https://www.masstransitmag.com/management/article/10220384/friending-transit,

400 Accessed May 8, 2019.

401 Gao, J.Z. 2017. Research on key technologies of bus priority signal control at intersection

402 based on bus GPS data., Dongnan University, Nanjing, China.

403 Gartner, N.H., Assmann, S.F., Lasaga, F., and Hou, D.L. 1991. A multiband approach to

404 arterial traffic signal optimization. Transportation Research Part B: Methodological,

405 25(1): 55-74.Hounsell, N.B., Shrestha, B.P., Mcleod, F.N., Palmer, S., Bowen, T., and

406 Head, J.R. 2007. Using global positioning system for bus priority in London: traffic

407 signals close to bus stops. Intelligent Transport Systems, 1(2):131-137.

408 Li, Z.L., Zhu, M.H., and Wang, B.J. 2015. Research on the arterial bus priority model

409 considering delays at the upstream and downstream intersections. Traffic Information and

410 Security, 33(5): 36-42.

411 Mcenakshy, V. 2005. Robust optimization model for bus priority under arterial progression.

412 University of Maryland, Maryland, Virginia.

413 Shaaban, K. and Ghanim, M. 2018. Evaluation of transit signal priority implementation for

414 bus transit along a major arterial using microsimulation. Procedia Computer Science, 130:

415 82-89.

416 Skabardonis, A. 2000. Control strategies for transit priority. Transportation Research Record

Page 17 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 19: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

18

417 1727, Journal of Transportation Research Board, Washington, D.C., 20-26.

418 Tien, T.L., Graham, C., Mark, W., Gruyter, C.D., and Kun, A. 2018. Coordinated transit

419 signal priority model considering stochastic bus arrival time. IEEE Transactions on

420 Intelligent Transportation Systems, 1-9.

421 Vincent, R.A., Cooper, B.R., and Wood, K. 1978. Bus-actuated signal control at isolated

422 intersections-simulation studies of bus priority. Transport and Road Research Laboratory,

423 Crowthorne, United Kingdom, 814, 21.

424 Wang, D.H., Tang, Y.H., and Chen, Q.G. 2015. Effective factors of interval time reliability of

425 bus station based on GPS data. Journal of Southeast University (Natural Science Edition).

426 45(02): 404-412.

427 Wang, H., Zhang, Z., Zhang, S., and Liu, X. 2014. Estimation of signalized intersections for

428 GPS data based on under-sampled traffic. Journal of Transportation Systems Engineering,

429 14(2): 51-56.

430 Wang, Y., Wang, T., and Shen, J. 2016. Active priority control method and evaluation of bus

431 signal based on overall delay of intersections. China Intelligent Transportation

432 Association, The Proceedings of the 11th China Intelligent Transportation Annual

433 Conference, 12: 448-459.

434 Wei, F. and Gurmu, Z. 2015. Dynamic travel time prediction models for buses using only

435 GPS data. International Journal of Transportation Science and Technology, 4(4):

436 353-366.

437 Woodard, D. , Nogin, G. , Koch, P. , Paul, Racz, D., Goldszmidt, M., and Horvitz, E. 2017.

438 Predicting travel time reliability using mobile phone GPS data. Transportation Research

439 Part C: Emerging Technologies, 75(C):30-44.

440 Xu, H.F., Li, K.P., and Zheng, M.M. 2008. Single point bus priority control strategy based on

441 logical rules. Journal of Chinese Highway, 5: 96-102.

442 Yu, Z.D., Liu, S.Q., Shen, W.C., Xu, J.M., and Lu, R.Q. 2015. Research on priority control

443 method of bus signal based on fuzzy control. Traffic Information and Security, 1: 35-40.

444 Zhen, Y. 2015. Research on the Technical Guarantee Technology of Urban Trunk Bus.

445 Nanjing: Dongnan University.

Page 18 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 20: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

19

446 Zhou, Z., Yang, J., Qi, and Y.,Cai, Y. 2018. Support vector machine and back propagation

447 neutral network approaches for trip mode prediction using mobile phone data. IET

448 Intelligent Transport Systems, 12(10):1220-1226.

449

450 FIGURES

451

452 Fig. 1 Logic of proposed analytical model for arterial bus priority

453 Fig. 2 Displacement time series plot of bus j in test k

454 Fig. 3 Procedures of single-cycle bus priority control logic

455 Fig. 4 Procedures of selecting bus priority timing plan for different traffic conditions

456 Fig. 5 Geometry of Jinshan Avenue and its intersections

457 Fig. 6 Existing signal timing for each intersection of Jinshan Avenue

458 Fig. 7 Traffic volume of the intersection at Jinshan Avenue at afternoon peak (pcu/h)

459 Fig. 8 Evaluation indicators for different common cycle lengths

460 Fig. 9 Evaluation indicators for different traffic flows

461 Fig. 10 Deviation of bus flow for different common cycle lengths

462

Page 19 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 21: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

1

1 LIST OF FIGURES

2

3 Fig. 1 Logic of proposed analytical model for arterial bus priority

4 Fig. 2 Displacement time series plot of bus j in test k

5 Fig. 3 Procedures of single-cycle bus priority control logic

6 Fig. 4 Procedures of selecting bus priority timing plan for different traffic conditions

7 Fig. 5 Geometry of Jinshan Avenue and its intersections

8 Fig. 6 Existing signal timing for each intersection of Jinshan Avenue

9 Fig. 7 Traffic volume of the intersection at Jinshan Avenue at afternoon peak (pcu/h)

10 Fig. 8 Evaluation indicators for different common cycle lengths

11 Fig. 9 Evaluation indicators for different traffic flows

12 Fig. 10 Deviation of bus flow for different common cycle lengths

Page 20 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 22: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

2

14

15 Fig. 1 Logic of proposed analytical model for arterial bus priority

Page 21 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 23: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

3

170Start point

End point

Bus j

Time(s)

Dis

tanc

e (m

)

18 Fig. 2 Displacement time series plot of bus j in test k

Page 22 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 24: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

4

20

21 Fig. 3 Procedures of single-cycle bus priority control logic

Page 23 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 25: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

5

23

24 Fig. 4 Procedures of selecting bus priority timing plan for different traffic conditions

25

Page 24 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 26: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

6

27

28 Fig. 5 Geometry of Jinshan Avenue and its intersections

Page 25 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 27: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

7

30

31

32

33 Fig. 6 Existing signal timing for each intersection of Jinshan Avenue

Page 26 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 28: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

8

34

35

36

37 Fig. 7 Traffic volume of the intersection at Jinshan Avenue at afternoon peak (pcu/h)

38

39

Page 27 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 29: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

9

850

900

950

1000

1050

1100

140 150 160 170 180

Ave

rage

Bus

Tra

vel T

ime(

s)

Common Cycle Length(s)

WEEWTWO-WAY

(a) Average bus travel time

10

11

12

13

14

15

16

17

18

140 150 160 170 180

Ave

rage

Num

ber o

f Bus

Sto

ps

Common Cycle Length(s)

WEEWTWO-WAY

(b) Average number of bus stops

400

450

500

550

600

650

140 150 160 170 180

Ave

rage

Bus

Del

ay(

s)

Common Cycle Length(s)

WEEWTWO-WAY

(c) Average bus delay

40 Fig. 8 Evaluation indicators for different common cycle lengths

Page 28 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 30: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

10

42

800

850

900

950

1000

1050

1100

1150

-20% -10% 0 +10% +20%

Ave

rage

Bus

Tra

vel T

ime(

s)

Change in Base Traffic Flow

140150160170180

(a) Average bus travel time

450

500

550

600

650

700

-20% -10% 0 +10% +20%

Ave

rage

Bus

Del

ay(

s)

Change in Base Traffic Flow

140150160170180

(b) Average bus delay

450

500

550

600

650

700

-20% -10% 0 +10% +20%

Ave

rage

Bus

Del

ay(

s)

Change in Base Traffic Flow

140150160170180

(c) Average number of bus stops

43 Fig. 9 Evaluation indicators for different traffic flows

Page 29 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 31: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

11

45

46

47

0

10

20

30

40

50

60

140 150 160 170 180

Dev

iatio

n of

Bus

Tra

ffic

Flo

w

Common Cycle Length(s)

Deviation of Bus Travel TimeDeviation of Bus DelayDeviation of Bus Stops

48 Fig. 10 Deviation of bus flow for different common cycle lengths

49

50

Page 30 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 32: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

1

1 LIST OF TABLES

2 Table 1 Sample of bus database from VISSIM simulation (Coordinate-EW)

3 Table 2 Descriptions of the fields of VISSIM simulation bus database

4 Table 3 Comparative analysis of two strategies

5 Table 4 Matching score table of each matching schedule

6 Table 5 Scores of evaluation indicators for combined plans of different common cycle lengths and

7 different traffic flows

8

Page 31 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 33: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

2

10

11 Table 1 Sample of bus database from VISSIM simulation (Coordinate-EW)

ID FLAG ELAPSEDTIME VEHICLEID POINT_X POINT_Y SPEED TYPE

244056 200816162045 223 1 6266.27947232137 1301.59777774379 17.5687219140412 300

244057 200816162045 224 1 6262.95800787531 1298.63276823481 14.2003606845484 300

244058 200816162045 225 1 6260.36252703294 1296.39639066689 10.3823676763503 300

244059 200816162045 226 1 6258.56286971133 1294.87336358465 6.56437466815222 300

244060 200816162045 227 1 6257.57135600081 1294.04110005584 2.74638165995412 300

244061 200816162045 228 1 6257.09739651233 1293.64467540518 1.69879662910241 300

244062 200816162045 229 1 6256.85594552073 1293.44307038223 0.564343159368555 300

244063 200816162045 230 1 6256.82516851697 1293.41737798768 0 300

244064 200816162045 231 1 6256.82516851697 1293.41737798768 0 300

Page 32 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 34: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

3

13

14 Table 2 Descriptions of the fields of VISSIM simulation bus database

Serial Number Fields Description

1 ID Identity2 FLAG Time (year and month)3 ELAPSEDTIME Current simulation seconds (s)4 VEHICLEID Vehicle number5 POINT_X X-coordinate6 POINT_Y Y-coordinate7 SPEED Current speed (km/h)8 TYPE Vehicle type number

15

Page 33 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 35: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

4

17 Table 3 Comparative analysis of two strategies

Evaluation IndicatorsGreen Wave

Coordinated Control Strategy

Single-Cycle Bus Priority Control Strategy

The average travel time (s) 984 963The average number of stops 16 15

The average delay (s) 556 531

Page 34 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 36: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

5

19

20 Table 4 Matching score table of each matching schedule

Score for Cycle LengthEvaluationIndicator 140 s 150 s 160 s 170 s 180 s

Average bus travel time 4 2 5 3 1Average number of bus stops 1 5 5 1 1

Average bus delay 3 2 5 4 1Total 8 9 15 8 3

21

Page 35 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering

Page 37: Modeling Arterial Signal Coordination for Bus Priority ... · Draft 1 Modeling Arterial Signal Coordination for Bus Priority Using Mobile-Phone GPS Data Yuanwen Lai,1 Xinying Xu,2

Draft

6

2324 Table 5 Scores of evaluation indicators for combined plans of different25 common cycle lengths and different traffic flows

26

27

28

29

3031

Change in Base Traffic FlowCommon Cycle Length (s) -20% -10% 0 +10% +20% Total

140 8 8 8 12 11 47150 7 8 8 13 7 43160 14 14 15 9 13 65170 6 7 8 6 14 41180 7 7 3 5 6 28

Page 36 of 36

https://mc06.manuscriptcentral.com/cjce-pubs

Canadian Journal of Civil Engineering