- Email: [email protected]
Countermeasure Strategies against Traffic Congestion on Motorways in Japan
Available online at www.sciencedirect.com
Procedia Social and Behavioral Sciences 16 (2011) 110–119
6th International Symposium on Highway Capacity and Quality of Service Stockholm, Sweden June 28 – July 1, 2011
Countermeasure Strategies against Traffic Congestion on Motorways in Japan Yoshiyasu Murashige* Expressway Research Institute, 1-4-1 Tadao Machida-shi 194-8508 Tokyo Japan
Abstract Due to excess demand, we encounter traffic congestion on motorways quite often at bottlenecks in the vicinity of metropolitan areas in Japan, mainly on holidays. The administrators of motorways have tackled this issue for more than four decades. At the beginning, they just tried to increase the number of lanes; however, widening of motorways requires huge construction costs and long periods. After the collapse of the bubble economy in the early 1990s, it became so difficult to broaden the existing motorways. To alleviate traffic congestion, they have introduced comparatively costless innovative attempts; however, those could not eradicate traffic congestion completely. This paper describes the advantageous effects of innovative attempts, methodologies and applied cases, such as traffic information system providing location of the traffic bottleneck to drivers to activate them to gain speed on the upgrade, using the variable message signs, or traffic demand management disseminating real-time traffic information (e.g., travel time, length of the queue) to drivers, resulting in mitigation of traffic congestion on motorways. In recent years, toll discount has been introduced on motorways as a pilot program; the situation of traffic congestion is getting worse. This paper also mentions a quick overview of the countermeasure strategies against traffic congestion and of the state of the congestion on motorways for a couple of decades. © 2011 Published by Elsevier Ltd. Keywords: Congestion, Countermeasures, Motorways, Information Provision System, Variable Message Signs;
1. Historical review of congestion on motorways in Japan After WWII, Japan’s postwar recovery accelerated the urbanization. Late 1960s the concentration of industry and population achieved a peek in the pacific belt zone between Tokyo and Osaka (cf. Figure 1). Today this area has 65 million population, covers 51%, and also accounts 57% of gross product of the nation, along the approx.540 km of
* Yoshiyasu Murashige. Tel.: +81-42-791-1673 Fax: +81-42-791-1099 E-mail address: [email protected]
1877-0428 © 2011 Published by Elsevier Ltd. doi:10.1016/j.sbspro.2011.04.434
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
111
the pacific coast, however, the number of total lanes of the motorway running through this area is just about four ( cf. table 1).
Tokyo-Osaka motorway
Figure 1: Motorway Network of Japan
Table 1: # of lanes of the motorway between Tokyo and Osaka # of lanes 4-lane 6-lane 7-lane 8-lane Total
Length (km) 421.1km 86.8km 20.6km 8.3km 536.8km
Percentage (%) 78.5% 16.2% 3.8% 1.5% 100.0%
This pacific belt zone between Tokyo and Osaka is the largest metropolitan area in the world in terms of population, economy, culture, and so on. Because the huge quantities of demand of traffic, generated by the socioeconomic activities of this area, have rushed into the four-lane Tokyo-Osaka motorway since its completion in 1969, the traffic congestion on the motorway has been a serious problem over a multiple generations.
112
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
Assuming that the average travel speed on the motorway (100km/h), the travel time between Tokyo and Osaka would be 322 min(5hrs 22min ), however, considering the delay by traffic congestion, the average travel time of all the vehicles traveling between Tokyo and Osaka at all times a year would be 342min (5hrs42). The delay of travel time, 20min, is almost 6% of the normal travel time between Tokyo and Osaka, therefore this delay shows that the motorway is heavily and chronically congested. Early 1980s, reconstruction work to increase the number of lanes of the motorways from 4-lane to 7,8-lane in the vicinity of Tokyo and Osaka metropolitan areas was planned and completed middle 1990s, however widening of motorways requires huge cost and ling period. The collapse of the Bubble Economy marked the beginning of a prolonged slump in the 1990s. With the sluggish economy in the back ground, the traffic demand continues to stagnate and motorway network construction and renovating existing motorways reduced the rate. Moreover, with the advent of the Koizumi administration (2001-2005), it tried to reform of the government's investment and loan program. As a starter, it embarked on privatizing the Japan Highway Public Corporation (JH). The reason why the JH should be targeted was simply because it was a one of the largest government’s inverters. There was no argument on whether decreasing the budget for investing motorways and privatizing a public sector would be better in the national economic point of view among the government. During the Koizumi administration, total budget for investing motorways had declined dramatically. It became much more difficult to broaden the existing motorways.
2. Background of cost-effective attempts Phenomena at the bottlenecks on Japanese motorways have been researched since early 1980s (Koshi et al., 1983; Koshi, 1986; Koshi et al., 1992). According to the results of above researches, the mechanism of congested queue occurrence and the relationship between the bottleneck capacity and drivers’ behavior become clear. Once congestion occurs at a bottleneck, such as a sag or a tunnel, the capacity would drop from pre-queue breakdown flow rate to queue discharge flow rate (QDFR). The capacity drop is mainly caused by insufficient acceleration around and downstream of a bottleneck, because drivers could hardly notice the bottleneck location or end of congested queue, and therefore it is difficult for them to accelerate sufficiently as expected (Koshi et al., 1992). Thus queue discharge flow rate (QDFR) could increase by providing information on location of a bottleneck or end of congested queue accurately to activate drivers’ acceleration behavior. Oguchi et al. (2001) showed traffic congestion tends to occur due to imbalanced lane usage, e.g. on 4-lane motorways, almost 60% of drovers use passing lane (right lane) at comparatively high speed, by contrast 40%of use driving lane (left lane), and proposed to introduce auxiliary lanes in order to reduce the usage of the passing lane (right lane). According to the proposal, at some bottlenecks auxiliary lanes have introduced and proved advantageous effect on increasing both pre-queue breakdown flow rate and queue discharge flow rate (QDFR). However, constructing a few kilometer long auxiliary lane costs correspondingly huge expenditure.
3. Activate drivers by providing information using the VMS display In 1992 the measure for mitigation of traffic congestion using the VMS was introduced, for the first time ever in Japan, on the two-way two-lane motorway section with a series of tunnels during holiday peak period. A truck equipped with VMS (cf. Photo 1, 2) displayed “Congestion queue ends after going through the tunnel” at the entrance of a tunnel.
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
Photo 1: VMS Display “Congested queue ends just ahead!”
Photo 2: VMS Display “Be careful not to slow down!”
113
Photo 3: VMS Display “Restore the speed please!”
As a result, the queue discharge flow rate (QDFR) at a bottleneck increased by 9%. The effect of the measure noted above was verified again at a bottleneck caused by a sag on the Tokyo-Osaka motorway in the vicinity of Yokohama in 2001. Using a truck equipped with VMS, former JH investigated effective VMS location and contents of displayed massages. Based on the results of the investigation, JH installed two overhead VMSs 300m apart around a bottleneck of a sag (cf. Photo 3). As for the queue discharge flow rate (QDFR), it increased by 7% from 4,720 veh/h (W/O VMS) to 5,045 veh/h (W/ VMS). Subsequently the measure for mitigation of traffic congestion using the VMS has become operative over the country.
4. Factor analyses of the queue discharge flows rate 4.1 Daytime and nighttime comparison We have 12 bottlenecks on Japanese 6-lane motorways where the measure noted above has been carried out. It is already known that the queue discharge flow rate (QDFR) is affected by the time spent in the congested queue (TIQ) (Koshi et al., 1992). Figure 2 illustrates a relationship between TIQ and the queue discharge flow rate (QDFR) (Nakatani et al., 2005). As shown in Figure 1, the queue discharge flow rate (QDFR) decreases with the TIQ when TIQ is shorter than 30 minutes and it maintains roughly constant when TIQ is longer than 30 minutes. In this paper, the effects of the measure for mitigation of traffic congestion using the VMS are analyzed under the consideration that the queue discharge flow rate (QDFR) is stable when TIQ is longer than 30 minutes. As for the incremental queue discharge flow rate (IQDFR), it seems higher during the daytime than during the nighttime, and IQDFR at a bottleneck, where the queue discharge flow rate (QDFR) is higher, yields smaller IQDFR than IQDFR at a bottleneck, where the queue discharge flow rate (QDFR) is lower (cf.Figure 3,4). 6000
Queue discharge rate (veh/5min)
5000
4000
y = -1.5x + 371
3000
y =328
2000
1000
0 0
10
20
30
40
Figure 2:TIQ and queue discharge flow rate
50
60
70
80
Time in the queue (min)
90
100
110
120
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
114
Figure2: TIQ and queue discharge flow rate
Figure3: Incremental queue discharge flow rate (veh/h) TIQ>30 min
Figure4: Incremental queue discharge flow rate (%) TIQ>30 min.
4.2 The effect of number of VMSs The effect of the number of VMSs on IQDFR is investigated at four bottlenecks. The results are as follows. The IQDFR at bottlenecks, where the number of the VMSs is two or more, tends to be larger than the IQDFR at bottlenecks, where the number of the VMS is only one (cf. Table 2) . Moreover, above mentioned tendency is more predominant during the daytime than during the nighttime. Table2: The effect of number of VMSs on IQDFR
.DQ(WVX 2XWERXQG .DQ(WVX 2XWERXQG .DQ(WVX ,QERXQG
%RWWOHQHFN/RFDWLRQ +LJDVKL
0DWVX\DPD
1XPEHURI906 V $YDUDJH ,4')5 SHUFHQWLOH 9HKK SHUFHQWLOH $YDUDJH ,4')5 SHUFHQWLOH 㸣 SHUFHQWLOH
RQH
'D\WLPH +DQD]RQR RQH WZRRUPRUH
'D\WLPH +DQD]RQR 1LJKWWLPH RQH WZRRUPRUH WZRRUPRUH
7RPHL ,QERXQG
4.3 The effect of contents of displayed messages of VMSs The effect of contents of displayed messages of VMSs on IQDFR is investigated at five bottlenecks. Comparing the IQDFR at bottlenecks, where the VMSs display the distance to the end of congested queue, such as “Congested queue ends just ahead!” or “300m ahead, congested queue ends!”, with that at bottlenecks, where
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
115
the VMSs don’t display the distance to the end of congested queue, there is no significant difference (cf. Table 3) . Table3: The effect of contents of displayed massages of VMSs on IQDFR
%RWWOHQHFN/RFDWLRQ
7RKRNX 2XWERXQG 7RKRNX 2XWERXQG .DQ(WVX 2XWERXQG 7RKRNX ,QERXQG 7RPHL ,QERXQG +LJDVKL 0DWVX\DPD
.XNL 'D\WLPH +DQ\X 'D\WLPH 'D\WLPH +DQ\X 1LJKWWLPH
Photo 5: VMS at 25.2kp on Tokyo-Osaka motorway in the vicinity of Yokohama
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
116
VMS (25.2kp)
VMS (26.2kp)
Restore the speed please!
Congested queue ends just ahead!
VMS (25.0kp) Blackout
Traffic Flow
Figure 5: Location and Contents of Displayed Massages of VMSs Table 4: Contents of Displayed Massages of VMSs
Case A C E
26.2KP (Upstream) Congested queue ends just ahead! Congested queue ends just ahead! Congested queue ends just ahead!
25.2KP (Bottom of the Sag)
25.0KP (Downstream) Be careful to drive! Congested
Restore the speed please! Restore the speed please!
Blackout
Restore the speed please! Restore the speed please!
ͤ Note that the case “C” has the very small number of samples Table 5: Comparison of QDFR 㻭㼢㼑㼞㼍㼓㼑㻔㼂㼑㼔㻛㼔㻕
'D\WLPH
㻯㼍㼟㼑
㻽㻰㻲㻾
㻝㻡㼜㼑㼞㼏㼑㼚㼠㼕㼘㼑㻔㼂㼑㼔㻛㼔㻕
㼕㼚㼏㼞㼑㼙㼑㼚㼠 㼕㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
㻽㻰㻲㻾
㼕㼚㼏㼞㼑㼙㼑㼚㼠 㼕㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
$
&
㻭㼢㼑㼞㼍㼓㼑㻔㼂㼑㼔㻛㼔㻕 㻽㻰㻲㻾
㻝㻡㼜㼑㼞㼏㼑㼚㼠㼕㼘㼑㻔㼂㼑㼔㻛㼔㻕
㼕㼚㼏㼞㼑㼙㼑㼚㼠 㼕㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
:2906
$
&
(
㻽㻰㻲㻾
㻏㻌㼛㼒 㼕㼚㼏㼞㼑㼙㼑㼚㼠 㻿㼍㼙㼜㼘㼑㼟 㼕㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
:2906
㻯㼍㼟㼑
1LJKWWLPH
㻤㻡㼜㼑㼞㼏㼑㼚㼠㼕㼘㼑㻔㼂㼑㼔㻛㼔㻕
㻽㻰㻲㻾
㼕㼚㼏㼞㼑㼙㼑㼚㼠 㼕㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
ͤ4')5㸸)ORZ㹰DWHLQPLQ 7,4!PLQ ࠉࠉ
㻤㻡㼜㼑㼞㼏㼑㼚㼠㼕㼘㼑㻔㼂㼑㼔㻛㼔㻕 㻽㻰㻲㻾
㻏㻌㼛㼒 㼕㼚㼏㼞㼑㼙㼑㼚㼠 㻿㼍㼙㼜㼘㼑㼟 㼕㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
㼘㼑㼢㼑㼘㻌㼛㼒㻌㼟㼕㼓㼚㼕㼒㼕㼏㼍㼚㼏㼑 㻌㼛㼒㻌㼍㻌㼠㻙㼠㼑㼟㼠㻌㻦㻌㻡㻑 㻔㼟㼕㼓㼚㼕㼒㼕㼏㼍㼚㼠㻕 㼃㻛㻻䋽㻭 㻭䋽㻯 㼘㼑㼢㼑㼘㻌㼛㼒㻌㼟㼕㼓㼚㼕㼒㼕㼏㼍㼚㼏㼑 㻌㼛㼒㻌㼍㻌㼠㻙㼠㼑㼟㼠㻌㻦㻌㻡㻑 㻔㼟㼕㼓㼚㼕㼒㼕㼏㼍㼚㼠㻕 㼃㻛㻻䋽㻱 㼃㻛㻻䋽㻭 㻭䋽㻱 㻯䋽㻱
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
Figure6: QDFR (veh/h) during Daytime (TIQ>30 min.)
Figure7: QDFR (veh/h) during Nighttime (TIQ>30 min.)
117
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
118
In the case “A” drivers encounter “Congested queue ends just ahead!” first, and “Restore the speed please!” then after that “Be careful to drive! Congested” This sequence of display confuses drivers. At first, drivers consider that the congested queue would end and try to accelerate, however, secondly they hesitate to do so, watching the last display. The case “A” lacks consistency of the massage to inform drivers how to behave properly. Therefore, the QDFR of the case “A” is considerably lower than that of the case “W/O VMS” (cf. Figure 6, 7, and 8). In the case “E” the massage “Restore the speed please!” is displayed repeatedly, and its QDFR is higher than that of the case “W/O VMS” in some degree (cf. Figure 7, 8, and 9). Figure8 indicates the IQDFR of the case “A”, “C” and “E” during daytime and nighttime compared with that of the case “W/O VMS”. 'D\WLPH
1LJKWWLPH
㻠㻜㻜
㻠㻜㻜
㻟㻜㻜
㻟㻜㻜
㻞㻜㻜
㻞㻜㻜
㻵㻽㻰㻲㻾䠄㼂㼑㼔㻛㼔䠅
㻵㻽㻰㻲㻾䠄㼂㼑㼔㻛㼔䠅
$YHUDJH
㻠㻣
㻝㻜㻜 㻜 㻙㻝㻜㻜 㻙㻞㻜㻜
㻝㻝㻥
㻝㻜㻜 㻜 㻙㻝㻜㻜
㻙㻢㻟
㻙㻞㻜㻜
㻙㻝㻠㻤
㻙㻟㻜㻜
㻙㻝㻢㻟
㻙㻟㻜㻜 㻭
㻯
㻭
㻯 㻯㼍㼟㼑
㻯㼍㼟㼑
㻱
Figure8: IQDFR (veh/h) compared with the case “W/O VMS” (TIQ>30 min.)
Table6: Contents of Displayed Massages of VMSs Case㻌 C㻌 D㻌
26.2KP (Upstream) Congested queue ends just ahead! Congested queue ends 1km ahead!㻌
25.2KP (Bottom of the Sag)
25.0KP (Downstream)
Restore the speed please!
Blackout
Restore the speed please!㻌
Blackout㻌
ͤ Note that the case “C” has the very small number of samples. Table7: Comparison of QDFR 㻭㼢㼑㼞㼍㼓㼑㻔㼂㼑㼔㻛㼔㻕
1LJKWWLPH
㻯㼍㼟㼑
㻽㻰㻲㻾
㻤㻡㼜㼑㼞㼏㼑㼚㼠㼕㼘㼑㻔㼂㼑㼔㻛㼔㻕
㻝㻡㼜㼑㼞㼏㼑㼚㼠㼕㼘㼑㻔㼂㼑㼔㻛㼔㻕
㻵㼚㼞㼑㼙㼑㼚㼠
㻵㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
㻽㻰㻲㻾
㻵㼚㼞㼑㼙㼑㼚㼠
㻵㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
㻽㻰㻲㻾
㻵㼚㼞㼑㼙㼑㼚㼠
㻵㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
㻏㻌㼛㼒 㼟㼍㼙㼜㼘㼑㼟
:2906
&
'
㼘㼑㼢㼑㼘㻌㼛㼒 㼟㼕㼓㼚㼕㼒㼕㼏㼍㼚㼏㼑 㻌㼛㼒㻌㼍㻌㼠㻙㼠㼑㼟㼠㻌㻦㻌㻡㻑 㻔㼟㼕㼓㼚㼕㼒㼕㼏㼍㼚㼠㻕 㼃㻛㻻䋽㻰 㻯䋽㻰
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
119
Figure9: QDFR (Veh/h) during Nighttime (TIQ>30 min.) In the case “D” compared with the case “W/O VMS” the exact distance to the end of congested queue, such as “Congested queue ends 1km ahead!” shows a modestly larger QDFR (cf. Figure 8). On the other hand, the case “C” indicates a smaller QDFR than that of the case “W/O VMS,” however, we should note that the case “C” has only thirteen samples. References Koshi, M., Iwasaki, M., and Ohkura, I. Some findings and an overview on vehicular flow characteristics. Proceedings of 8th International Symposium on transportation and Traffic flow Theory, 1983, pp.403-426 Koshi, M., ‘Capacity of motorway bottlenecks’, Journal of JSCE, Vol.371 / IV-5, 1986, pp. 1-7 Koshi, M., Kuwahara, M. and Akahane, H., ‘Capacity of sags and tunnels on Japanese motorways’, ITE Journal. Vol.62 No.5, 1992, pp. 17-22 Oguchi, T., Kuwahara, M., Akabane, H., and Watanabe, A., Auxiliary lane configuration controlling traffic lane distribution at the upstream of a bottleneck, Japan Society of Traffic Engineers 36(1), 2001. Pp.56-69 (in Japanese) Nakatani, Minagata, T., Sato, H. and Ichikawa, M., ‘Evaluation on effect of provision of speed increase advice message by LED sign in consideration of time spent in queue’, Proc. of the 25th JSTE Annual Conference, 2005, pp. 173-176 Saito, Y., Xing, J., Nonaka, Y., Ishida, T. and Uchiyama, H. (2005) ‘An experimental study on mitigation of expressway traffic congestion with LED information board’, Proc. of the EASTS (CD-ROM), Vol.5, pp. 919-928 J. Xing, H. Takahashi and T. Takeuchi, “Increasing bottleneck capacity through provision of bottleneck location information”, Proceedings of the 11th World Conference on Transportation Research (CD-ROM), June 2007 Xing, J. and Koshi, M., ‘A study on the bottleneck phenomena and car-following behavior on sags of motorways’, Journal of JSCE, Vol.506 / IV-26, 1995, pp. 45-55 Nonaka Y., Kasai M., Ishida T. (2004): Development and Adaptation of a system to obtain data of car behavior on an interurban expressway, Proceedings 24th JSTE conference, 25-28 Murashige,Y., ‘Mitigation of Traffic Congestion on Motorways by Increasing Queue Discharge Flow Rate Using the VMSs’, 16th ITS World Congress Stockholm, 2009 proceedings (CD-ROM)
Procedia Social and Behavioral Sciences 16 (2011) 110–119
6th International Symposium on Highway Capacity and Quality of Service Stockholm, Sweden June 28 – July 1, 2011
Countermeasure Strategies against Traffic Congestion on Motorways in Japan Yoshiyasu Murashige* Expressway Research Institute, 1-4-1 Tadao Machida-shi 194-8508 Tokyo Japan
Abstract Due to excess demand, we encounter traffic congestion on motorways quite often at bottlenecks in the vicinity of metropolitan areas in Japan, mainly on holidays. The administrators of motorways have tackled this issue for more than four decades. At the beginning, they just tried to increase the number of lanes; however, widening of motorways requires huge construction costs and long periods. After the collapse of the bubble economy in the early 1990s, it became so difficult to broaden the existing motorways. To alleviate traffic congestion, they have introduced comparatively costless innovative attempts; however, those could not eradicate traffic congestion completely. This paper describes the advantageous effects of innovative attempts, methodologies and applied cases, such as traffic information system providing location of the traffic bottleneck to drivers to activate them to gain speed on the upgrade, using the variable message signs, or traffic demand management disseminating real-time traffic information (e.g., travel time, length of the queue) to drivers, resulting in mitigation of traffic congestion on motorways. In recent years, toll discount has been introduced on motorways as a pilot program; the situation of traffic congestion is getting worse. This paper also mentions a quick overview of the countermeasure strategies against traffic congestion and of the state of the congestion on motorways for a couple of decades. © 2011 Published by Elsevier Ltd. Keywords: Congestion, Countermeasures, Motorways, Information Provision System, Variable Message Signs;
1. Historical review of congestion on motorways in Japan After WWII, Japan’s postwar recovery accelerated the urbanization. Late 1960s the concentration of industry and population achieved a peek in the pacific belt zone between Tokyo and Osaka (cf. Figure 1). Today this area has 65 million population, covers 51%, and also accounts 57% of gross product of the nation, along the approx.540 km of
* Yoshiyasu Murashige. Tel.: +81-42-791-1673 Fax: +81-42-791-1099 E-mail address: [email protected]
1877-0428 © 2011 Published by Elsevier Ltd. doi:10.1016/j.sbspro.2011.04.434
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
111
the pacific coast, however, the number of total lanes of the motorway running through this area is just about four ( cf. table 1).
Tokyo-Osaka motorway
Figure 1: Motorway Network of Japan
Table 1: # of lanes of the motorway between Tokyo and Osaka # of lanes 4-lane 6-lane 7-lane 8-lane Total
Length (km) 421.1km 86.8km 20.6km 8.3km 536.8km
Percentage (%) 78.5% 16.2% 3.8% 1.5% 100.0%
This pacific belt zone between Tokyo and Osaka is the largest metropolitan area in the world in terms of population, economy, culture, and so on. Because the huge quantities of demand of traffic, generated by the socioeconomic activities of this area, have rushed into the four-lane Tokyo-Osaka motorway since its completion in 1969, the traffic congestion on the motorway has been a serious problem over a multiple generations.
112
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
Assuming that the average travel speed on the motorway (100km/h), the travel time between Tokyo and Osaka would be 322 min(5hrs 22min ), however, considering the delay by traffic congestion, the average travel time of all the vehicles traveling between Tokyo and Osaka at all times a year would be 342min (5hrs42). The delay of travel time, 20min, is almost 6% of the normal travel time between Tokyo and Osaka, therefore this delay shows that the motorway is heavily and chronically congested. Early 1980s, reconstruction work to increase the number of lanes of the motorways from 4-lane to 7,8-lane in the vicinity of Tokyo and Osaka metropolitan areas was planned and completed middle 1990s, however widening of motorways requires huge cost and ling period. The collapse of the Bubble Economy marked the beginning of a prolonged slump in the 1990s. With the sluggish economy in the back ground, the traffic demand continues to stagnate and motorway network construction and renovating existing motorways reduced the rate. Moreover, with the advent of the Koizumi administration (2001-2005), it tried to reform of the government's investment and loan program. As a starter, it embarked on privatizing the Japan Highway Public Corporation (JH). The reason why the JH should be targeted was simply because it was a one of the largest government’s inverters. There was no argument on whether decreasing the budget for investing motorways and privatizing a public sector would be better in the national economic point of view among the government. During the Koizumi administration, total budget for investing motorways had declined dramatically. It became much more difficult to broaden the existing motorways.
2. Background of cost-effective attempts Phenomena at the bottlenecks on Japanese motorways have been researched since early 1980s (Koshi et al., 1983; Koshi, 1986; Koshi et al., 1992). According to the results of above researches, the mechanism of congested queue occurrence and the relationship between the bottleneck capacity and drivers’ behavior become clear. Once congestion occurs at a bottleneck, such as a sag or a tunnel, the capacity would drop from pre-queue breakdown flow rate to queue discharge flow rate (QDFR). The capacity drop is mainly caused by insufficient acceleration around and downstream of a bottleneck, because drivers could hardly notice the bottleneck location or end of congested queue, and therefore it is difficult for them to accelerate sufficiently as expected (Koshi et al., 1992). Thus queue discharge flow rate (QDFR) could increase by providing information on location of a bottleneck or end of congested queue accurately to activate drivers’ acceleration behavior. Oguchi et al. (2001) showed traffic congestion tends to occur due to imbalanced lane usage, e.g. on 4-lane motorways, almost 60% of drovers use passing lane (right lane) at comparatively high speed, by contrast 40%of use driving lane (left lane), and proposed to introduce auxiliary lanes in order to reduce the usage of the passing lane (right lane). According to the proposal, at some bottlenecks auxiliary lanes have introduced and proved advantageous effect on increasing both pre-queue breakdown flow rate and queue discharge flow rate (QDFR). However, constructing a few kilometer long auxiliary lane costs correspondingly huge expenditure.
3. Activate drivers by providing information using the VMS display In 1992 the measure for mitigation of traffic congestion using the VMS was introduced, for the first time ever in Japan, on the two-way two-lane motorway section with a series of tunnels during holiday peak period. A truck equipped with VMS (cf. Photo 1, 2) displayed “Congestion queue ends after going through the tunnel” at the entrance of a tunnel.
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
Photo 1: VMS Display “Congested queue ends just ahead!”
Photo 2: VMS Display “Be careful not to slow down!”
113
Photo 3: VMS Display “Restore the speed please!”
As a result, the queue discharge flow rate (QDFR) at a bottleneck increased by 9%. The effect of the measure noted above was verified again at a bottleneck caused by a sag on the Tokyo-Osaka motorway in the vicinity of Yokohama in 2001. Using a truck equipped with VMS, former JH investigated effective VMS location and contents of displayed massages. Based on the results of the investigation, JH installed two overhead VMSs 300m apart around a bottleneck of a sag (cf. Photo 3). As for the queue discharge flow rate (QDFR), it increased by 7% from 4,720 veh/h (W/O VMS) to 5,045 veh/h (W/ VMS). Subsequently the measure for mitigation of traffic congestion using the VMS has become operative over the country.
4. Factor analyses of the queue discharge flows rate 4.1 Daytime and nighttime comparison We have 12 bottlenecks on Japanese 6-lane motorways where the measure noted above has been carried out. It is already known that the queue discharge flow rate (QDFR) is affected by the time spent in the congested queue (TIQ) (Koshi et al., 1992). Figure 2 illustrates a relationship between TIQ and the queue discharge flow rate (QDFR) (Nakatani et al., 2005). As shown in Figure 1, the queue discharge flow rate (QDFR) decreases with the TIQ when TIQ is shorter than 30 minutes and it maintains roughly constant when TIQ is longer than 30 minutes. In this paper, the effects of the measure for mitigation of traffic congestion using the VMS are analyzed under the consideration that the queue discharge flow rate (QDFR) is stable when TIQ is longer than 30 minutes. As for the incremental queue discharge flow rate (IQDFR), it seems higher during the daytime than during the nighttime, and IQDFR at a bottleneck, where the queue discharge flow rate (QDFR) is higher, yields smaller IQDFR than IQDFR at a bottleneck, where the queue discharge flow rate (QDFR) is lower (cf.Figure 3,4). 6000
Queue discharge rate (veh/5min)
5000
4000
y = -1.5x + 371
3000
y =328
2000
1000
0 0
10
20
30
40
Figure 2:TIQ and queue discharge flow rate
50
60
70
80
Time in the queue (min)
90
100
110
120
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
114
Figure2: TIQ and queue discharge flow rate
Figure3: Incremental queue discharge flow rate (veh/h) TIQ>30 min
Figure4: Incremental queue discharge flow rate (%) TIQ>30 min.
4.2 The effect of number of VMSs The effect of the number of VMSs on IQDFR is investigated at four bottlenecks. The results are as follows. The IQDFR at bottlenecks, where the number of the VMSs is two or more, tends to be larger than the IQDFR at bottlenecks, where the number of the VMS is only one (cf. Table 2) . Moreover, above mentioned tendency is more predominant during the daytime than during the nighttime. Table2: The effect of number of VMSs on IQDFR
.DQ(WVX 2XWERXQG .DQ(WVX 2XWERXQG .DQ(WVX ,QERXQG
%RWWOHQHFN/RFDWLRQ +LJDVKL
0DWVX\DPD
1XPEHURI906 V $YDUDJH ,4')5 SHUFHQWLOH 9HKK SHUFHQWLOH $YDUDJH ,4')5 SHUFHQWLOH 㸣 SHUFHQWLOH
RQH
'D\WLPH +DQD]RQR RQH WZRRUPRUH
'D\WLPH +DQD]RQR 1LJKWWLPH RQH WZRRUPRUH WZRRUPRUH
7RPHL ,QERXQG
4.3 The effect of contents of displayed messages of VMSs The effect of contents of displayed messages of VMSs on IQDFR is investigated at five bottlenecks. Comparing the IQDFR at bottlenecks, where the VMSs display the distance to the end of congested queue, such as “Congested queue ends just ahead!” or “300m ahead, congested queue ends!”, with that at bottlenecks, where
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
115
the VMSs don’t display the distance to the end of congested queue, there is no significant difference (cf. Table 3) . Table3: The effect of contents of displayed massages of VMSs on IQDFR
%RWWOHQHFN/RFDWLRQ
7RKRNX 2XWERXQG 7RKRNX 2XWERXQG .DQ(WVX 2XWERXQG 7RKRNX ,QERXQG 7RPHL ,QERXQG +LJDVKL 0DWVX\DPD
.XNL 'D\WLPH +DQ\X 'D\WLPH 'D\WLPH +DQ\X 1LJKWWLPH
Photo 5: VMS at 25.2kp on Tokyo-Osaka motorway in the vicinity of Yokohama
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
116
VMS (25.2kp)
VMS (26.2kp)
Restore the speed please!
Congested queue ends just ahead!
VMS (25.0kp) Blackout
Traffic Flow
Figure 5: Location and Contents of Displayed Massages of VMSs Table 4: Contents of Displayed Massages of VMSs
Case A C E
26.2KP (Upstream) Congested queue ends just ahead! Congested queue ends just ahead! Congested queue ends just ahead!
25.2KP (Bottom of the Sag)
25.0KP (Downstream) Be careful to drive! Congested
Restore the speed please! Restore the speed please!
Blackout
Restore the speed please! Restore the speed please!
ͤ Note that the case “C” has the very small number of samples Table 5: Comparison of QDFR 㻭㼢㼑㼞㼍㼓㼑㻔㼂㼑㼔㻛㼔㻕
'D\WLPH
㻯㼍㼟㼑
㻽㻰㻲㻾
㻝㻡㼜㼑㼞㼏㼑㼚㼠㼕㼘㼑㻔㼂㼑㼔㻛㼔㻕
㼕㼚㼏㼞㼑㼙㼑㼚㼠 㼕㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
㻽㻰㻲㻾
㼕㼚㼏㼞㼑㼙㼑㼚㼠 㼕㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
$
&
㻭㼢㼑㼞㼍㼓㼑㻔㼂㼑㼔㻛㼔㻕 㻽㻰㻲㻾
㻝㻡㼜㼑㼞㼏㼑㼚㼠㼕㼘㼑㻔㼂㼑㼔㻛㼔㻕
㼕㼚㼏㼞㼑㼙㼑㼚㼠 㼕㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
:2906
$
&
(
㻽㻰㻲㻾
㻏㻌㼛㼒 㼕㼚㼏㼞㼑㼙㼑㼚㼠 㻿㼍㼙㼜㼘㼑㼟 㼕㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
:2906
㻯㼍㼟㼑
1LJKWWLPH
㻤㻡㼜㼑㼞㼏㼑㼚㼠㼕㼘㼑㻔㼂㼑㼔㻛㼔㻕
㻽㻰㻲㻾
㼕㼚㼏㼞㼑㼙㼑㼚㼠 㼕㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
ͤ4')5㸸)ORZ㹰DWHLQPLQ 7,4!PLQ ࠉࠉ
㻤㻡㼜㼑㼞㼏㼑㼚㼠㼕㼘㼑㻔㼂㼑㼔㻛㼔㻕 㻽㻰㻲㻾
㻏㻌㼛㼒 㼕㼚㼏㼞㼑㼙㼑㼚㼠 㻿㼍㼙㼜㼘㼑㼟 㼕㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
㼘㼑㼢㼑㼘㻌㼛㼒㻌㼟㼕㼓㼚㼕㼒㼕㼏㼍㼚㼏㼑 㻌㼛㼒㻌㼍㻌㼠㻙㼠㼑㼟㼠㻌㻦㻌㻡㻑 㻔㼟㼕㼓㼚㼕㼒㼕㼏㼍㼚㼠㻕 㼃㻛㻻䋽㻭 㻭䋽㻯 㼘㼑㼢㼑㼘㻌㼛㼒㻌㼟㼕㼓㼚㼕㼒㼕㼏㼍㼚㼏㼑 㻌㼛㼒㻌㼍㻌㼠㻙㼠㼑㼟㼠㻌㻦㻌㻡㻑 㻔㼟㼕㼓㼚㼕㼒㼕㼏㼍㼚㼠㻕 㼃㻛㻻䋽㻱 㼃㻛㻻䋽㻭 㻭䋽㻱 㻯䋽㻱
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
Figure6: QDFR (veh/h) during Daytime (TIQ>30 min.)
Figure7: QDFR (veh/h) during Nighttime (TIQ>30 min.)
117
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
118
In the case “A” drivers encounter “Congested queue ends just ahead!” first, and “Restore the speed please!” then after that “Be careful to drive! Congested” This sequence of display confuses drivers. At first, drivers consider that the congested queue would end and try to accelerate, however, secondly they hesitate to do so, watching the last display. The case “A” lacks consistency of the massage to inform drivers how to behave properly. Therefore, the QDFR of the case “A” is considerably lower than that of the case “W/O VMS” (cf. Figure 6, 7, and 8). In the case “E” the massage “Restore the speed please!” is displayed repeatedly, and its QDFR is higher than that of the case “W/O VMS” in some degree (cf. Figure 7, 8, and 9). Figure8 indicates the IQDFR of the case “A”, “C” and “E” during daytime and nighttime compared with that of the case “W/O VMS”. 'D\WLPH
1LJKWWLPH
㻠㻜㻜
㻠㻜㻜
㻟㻜㻜
㻟㻜㻜
㻞㻜㻜
㻞㻜㻜
㻵㻽㻰㻲㻾䠄㼂㼑㼔㻛㼔䠅
㻵㻽㻰㻲㻾䠄㼂㼑㼔㻛㼔䠅
$YHUDJH
㻠㻣
㻝㻜㻜 㻜 㻙㻝㻜㻜 㻙㻞㻜㻜
㻝㻝㻥
㻝㻜㻜 㻜 㻙㻝㻜㻜
㻙㻢㻟
㻙㻞㻜㻜
㻙㻝㻠㻤
㻙㻟㻜㻜
㻙㻝㻢㻟
㻙㻟㻜㻜 㻭
㻯
㻭
㻯 㻯㼍㼟㼑
㻯㼍㼟㼑
㻱
Figure8: IQDFR (veh/h) compared with the case “W/O VMS” (TIQ>30 min.)
Table6: Contents of Displayed Massages of VMSs Case㻌 C㻌 D㻌
26.2KP (Upstream) Congested queue ends just ahead! Congested queue ends 1km ahead!㻌
25.2KP (Bottom of the Sag)
25.0KP (Downstream)
Restore the speed please!
Blackout
Restore the speed please!㻌
Blackout㻌
ͤ Note that the case “C” has the very small number of samples. Table7: Comparison of QDFR 㻭㼢㼑㼞㼍㼓㼑㻔㼂㼑㼔㻛㼔㻕
1LJKWWLPH
㻯㼍㼟㼑
㻽㻰㻲㻾
㻤㻡㼜㼑㼞㼏㼑㼚㼠㼕㼘㼑㻔㼂㼑㼔㻛㼔㻕
㻝㻡㼜㼑㼞㼏㼑㼚㼠㼕㼘㼑㻔㼂㼑㼔㻛㼔㻕
㻵㼚㼞㼑㼙㼑㼚㼠
㻵㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
㻽㻰㻲㻾
㻵㼚㼞㼑㼙㼑㼚㼠
㻵㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
㻽㻰㻲㻾
㻵㼚㼞㼑㼙㼑㼚㼠
㻵㼚㼏㼞㼑㼙㼑㼚㼠 䠄䠂䠅
㻏㻌㼛㼒 㼟㼍㼙㼜㼘㼑㼟
:2906
&
'
㼘㼑㼢㼑㼘㻌㼛㼒 㼟㼕㼓㼚㼕㼒㼕㼏㼍㼚㼏㼑 㻌㼛㼒㻌㼍㻌㼠㻙㼠㼑㼟㼠㻌㻦㻌㻡㻑 㻔㼟㼕㼓㼚㼕㼒㼕㼏㼍㼚㼠㻕 㼃㻛㻻䋽㻰 㻯䋽㻰
Yoshiyasu Murashige / Procedia Social and Behavioral Sciences 16 (2011) 110–119
119
Figure9: QDFR (Veh/h) during Nighttime (TIQ>30 min.) In the case “D” compared with the case “W/O VMS” the exact distance to the end of congested queue, such as “Congested queue ends 1km ahead!” shows a modestly larger QDFR (cf. Figure 8). On the other hand, the case “C” indicates a smaller QDFR than that of the case “W/O VMS,” however, we should note that the case “C” has only thirteen samples. References Koshi, M., Iwasaki, M., and Ohkura, I. Some findings and an overview on vehicular flow characteristics. Proceedings of 8th International Symposium on transportation and Traffic flow Theory, 1983, pp.403-426 Koshi, M., ‘Capacity of motorway bottlenecks’, Journal of JSCE, Vol.371 / IV-5, 1986, pp. 1-7 Koshi, M., Kuwahara, M. and Akahane, H., ‘Capacity of sags and tunnels on Japanese motorways’, ITE Journal. Vol.62 No.5, 1992, pp. 17-22 Oguchi, T., Kuwahara, M., Akabane, H., and Watanabe, A., Auxiliary lane configuration controlling traffic lane distribution at the upstream of a bottleneck, Japan Society of Traffic Engineers 36(1), 2001. Pp.56-69 (in Japanese) Nakatani, Minagata, T., Sato, H. and Ichikawa, M., ‘Evaluation on effect of provision of speed increase advice message by LED sign in consideration of time spent in queue’, Proc. of the 25th JSTE Annual Conference, 2005, pp. 173-176 Saito, Y., Xing, J., Nonaka, Y., Ishida, T. and Uchiyama, H. (2005) ‘An experimental study on mitigation of expressway traffic congestion with LED information board’, Proc. of the EASTS (CD-ROM), Vol.5, pp. 919-928 J. Xing, H. Takahashi and T. Takeuchi, “Increasing bottleneck capacity through provision of bottleneck location information”, Proceedings of the 11th World Conference on Transportation Research (CD-ROM), June 2007 Xing, J. and Koshi, M., ‘A study on the bottleneck phenomena and car-following behavior on sags of motorways’, Journal of JSCE, Vol.506 / IV-26, 1995, pp. 45-55 Nonaka Y., Kasai M., Ishida T. (2004): Development and Adaptation of a system to obtain data of car behavior on an interurban expressway, Proceedings 24th JSTE conference, 25-28 Murashige,Y., ‘Mitigation of Traffic Congestion on Motorways by Increasing Queue Discharge Flow Rate Using the VMSs’, 16th ITS World Congress Stockholm, 2009 proceedings (CD-ROM)