Investigating the effects of gradual deployment of market penetration rates (MPR) of connected vehicles on delay time and fuel consumption

Abstract

Purpose

This research paper aims to investigate the effects of gradual deployment of market penetration rates (MPR) of connected vehicles (MPR of CVs) on delay time and fuel consumption.

Design/methodology/approach

A real-world origin-destination demand matrix survey was conducted in Boston, MA to identify the number of peak hour passing vehicles in the case study.

Findings

The results showed that as the number of CVs (MPR) in the network increases, the total delay time decreases by an average of 14% and the fuel consumption decreases by an average of 56%, respectively, from scenarios 3 to 15 compared to scenario 2.

Research limitations/implications

The first limitation of this study was considering a small network. The considered network shows a small part of the case study.

Originality/value

This study can be a milestone for future research regarding gradual deployment of CVs’ effects on transport networks. Efficient policy(s) may define based on the results of this network for Brockton transport network.

Citation

Ansariyar, A. and Tahmasebi, M. (2022), "Investigating the effects of gradual deployment of market penetration rates (MPR) of connected vehicles on delay time and fuel consumption", Journal of Intelligent and Connected Vehicles, Vol. 5 No. 3, pp. 188-198. https://doi.org/10.1108/JICV-12-2021-0018

1. Introduction

With the increasing number of vehicles and limited road network, the urban traffic congestion and the probability of incidents happening is growing at an alarming rate. The application of intelligent technology (IT) in the area of transportation and traffic engineering has improved significantly over the recent few years. Providing the connection technology between vehicles (V2V) and vehicles to infrastructure (V2I) is one of the most effective achievements in this area. The recent development in communication technologies facilitates the deployment of connected vehicles (CVs), which has been considered to significantly improve traffic safety and mobility in the transportation network. Moreover, the emergence of CV technology provided a light at the end of the tunnel, solving traffic congestion problems. Due to CVs’ benefits especially at congestion time (dynamic routing guidance and real-time connection), these new types of vehicles will spread in the not too distant future. To illustrate and assess the impression of the CV on transport networks, we developed a hybrid microsimulation model based on vehicle to everything (V2X) module in AIMSUN. The proposed hybrid model is able to simulate the driving behavior of CV guidance with various penetration rates of roadside units (RSU) and CVs.

We took advantage of the V2X module in AIMSUN 8.4.0 to efficiently simulate CVs impression. This module provides a heuristic architecture consisting of CVs, different types of messages to send and receive among vehicles and infrastructure, different RSUs and traffic management centers (TMC) to integrate the delivered messages. A real-world case study at Brockton, MA, was considered, and an origin-destination (OD) demand survey was conducted to determine the current traffic condition of the case study. Then, 15 scenarios consisting of different combinations of vehicle types (non-CV car, non-CV bus, CV-car and CV-bus) and different market penetration rates (MPR) of CVs were developed. The findings from this research will provide insight into the impacts of the gradual deployment of CVs and RSU on mobility and equity, which helps planners develop public policies on advanced vehicle promotion and regulations.

The remaining part of the paper aims to review recently published papers in terms of CAVs' dynamic rerouting. Hereupon, to identify the impact of CV route guidance technology in improving mobility and safety, a complete review of CV simulation studies through microsimulation software will be presented in Section 2. The methodology of this research will be organized in Section 3. Description of the V2X module and its architecture in AIMSUN software will be presented in Section 4. Explanation of the case study and detail of scenarios will be presented in Section 5. In Section 6, we will investigate the impression of MPR of CVs on the delay time and fuel consumption. Finally, we will draw conclusions on this research and discuss further developments and potentials for future research. It is worth noting that the references will be presented in the references section.

2. Literature review

Considering the rapid boom in information technology and people’s increasing dependence on mobile data, automotive manufacturers have started equipping vehicles with wireless communication capabilities, manufacturing what is commonly known as CVs, and autonomous systems to assist drivers with certain driving tasks. The communication devices in the vehicles make information available to either the vehicle or the driver, allowing them to interact with parts of the road infrastructure as well as other users on the road (Johnson, 2017). A number of technologies help achieve connectivity on the road and will assist drivers to accurately navigate the transport network. Two types of main connections are vehicle-to-vehicle (V2V) and vehicle-to-infrastructure or vice versa (V2I or I2V). The information provided to cars is expected to improve drivers’ efficiency, response and comfort while enhancing safety and mobility. As we mentioned earlier, this study will investigate the effects of increasing the MPR of CVs and its role in decreasing traffic jams or congestion on the urban transport network.

In the previous studies, multiple equilibrium behaviors along with sensitivity analysis were applied to estimate the impact of different MPR of CVs in various environments (Shladover et al., 2007; Jung et al., 2010; Rahman et al., 2021). Since the last decade, micro-simulation studies have been conducted in different environments, weather, road network and traffic level to evaluate the impact and performance of CVs. Meanwhile, software-based micro-simulation has been used to evaluate the feasibility and effectiveness of various approaches to improving mobility in advanced transportation systems. PARAMICS was used to set adverse weather conditions along with incidents in the simulation environment and identified the positive impact of CV deployment in reducing network travel time within moderate to a high level of traffic congestion (Gaspar and Nemeth, 2014). Whereas the safety aspect associated with different levels of market penetration of CVs was studied in 2016, and the result showed that less than 40% of market penetration of CVs provided a safer traffic network in special road conditions like a work zone to maintain the standard mobility (Chin et al., 2015). Similarly, micro-simulation was used to test an application developed for lane-speed monitoring based on V2V communication to identify in-lane real-time vehicle status (Xiao et al., 2017). The results demonstrated the effectiveness of CVs in improving all three aspects and provided quantitative results on how the MPR proportionally affects the performance of the traffic network. Other than the impact of CV technologies in the traffic system, this study also considered information updating time intervals.

Several types of research have been performed to identify and evaluate the positive impact of CV technologies on enhancing mobility and safety, but none has concentrated on quantifying these benefits under a realistic environment (Rahman et al., 2021). The effect of MPR and connection range (CR) of CVs in a traffic network were studied by developing both analytical and simulation models. The result showed that critical MP is sensitive to CR and vice versa, and reasonable MP and CR of CVs can reduce average travel time by 20% (Talebpour and Mahmassani., 2016).

To identify the impact of CV route guidance technology in improving mobility and safety, an urban traffic network was designed by Talebpour et al. (2011) in a micro-simulator. Mobility performance was measured by the average trip time (ATT) and average vehicle trip speed along with s surrogate measure – the time-to-collision involving incident rate (TTC-IR) – was used to assess safety. Observation of CV ATTs showed that it decreased with the increased market penetration of CVs from 0% to 50% and the opposite for non-CVs. However, as the time interval can be updated due to the research requirements, the CV-ATT may be increased by changing the time interval (s) of the traffic assignment. Hereupon, the CV-ATT was increased in this study in different MPR of CVs, even MPRs more than 50% which concluded that the updating interval lesser the ATTs of CVs. Considering the results of three factors, the study suggested that traffic mobility might reduce traffic safety with an increasing penetration level of CVs at different time intervals. Authors (Van Arem et al., 2006) used VISSIM micro-simulator for various route guidance strategies of CVs and considered factors like MPR of CVs, congestion levels, updating intervals of route guidance information and drivers’ compliance rates (Jaworski et al., 2012). Traffic congestion can be improved by 7%–12% depending on different CV MPR as identified while evaluating variable speed limit in the micro-simulation environment (Goodall et al., 2013). On the other hand, a micro-simulator was used to evaluate the potential benefit of CVs in identifying spillbacks and reducing system-wide cycle time (Gora, 2017). Besides that, a micro-simulator was applied to collect signal location and timing information and transmit via V2I communication with the suggested speed to reduce stopping time in traffic signals (Zhang and Cassandras, 2018).

Connected and automated vehicles (CAVs) are expected to improve both traffic safety and efficiency by reducing human driver errors. A widespread and quick propagation of information related to traffic incidents is crucial for managing them, as such incidents lead to secondary incidents, which account for 20% of all incidents (FHWA). If a vehicle is disabled due to an incident or mechanical failure, that information must be sent quickly to the approaching traffic (Cao et al., 2021). They used the average propagation distance of wireless messages to assess the performance of CAVs or cooperative vehicle systems that relied on traffic density and vehicle MPRs. With an increase in traffic density and MPR, the distance of message propagation was increased. This message propagation distance also increased quickly as the ratio between mean separation among the vehicles and communication range increased. The performance of inter-vehicle communications (IVC) was studied by Lochert et al. (2005). They used an NS-2 communication network simulator. They found that the average maximum distance of information propagation increases when there is an increase in the transmission range, as low traffic density and a shorter transmission range negatively impact the message propagation in IVC over several vehicles. It is worth noting that they considered equipped vehicle market penetration of only one level, i.e. 10%. Last but not least, managing traffic with a combination of CVs and non-CV cars would lead the planners to a complex problem. To simplify this concern, a strategy called platooning that leads by a CAV driven by a driver was introduced. By increasing the MPR by more than 60%, the benefit of driving CAVs would improve significantly (Yao et al., 2019).

After reviewing the aforementioned studies, it is obvious that the majority of studies did not consider the changes of delay time with different MPR of CVs. This study will try to investigate the impression of gradual deployment of CVs on the transport network and their role in better navigation and dynamic rerouting. This study will use the V2X module in micro-simulation software (AIMSUN) to model CVs in a fairly realistic environment, and then, it will evaluate the improvement of total delay time (sec/km) with gradual increasing the MPR of CVs on a real-world urban transport network.

3. Research methodology

To model the connectivity and rerouting of CVs, a new simulation environment was developed in AIMSUN software. Vehicles were divided into connected and non-connected. CVs are able to save and broadcast their travel time information observations. Leading CVs are able to observe and save link travel time information on time basis; this information will be disseminated to other CVs or RSUs if within the range of dedicated short-range communication (DSRC). The maximum range of DSRC communication between the TMC, RSU and CVs was specified 300 m (984 ft) through investigating the different types of RSUs available in the market. This number (300 m) was determined through investigating the different types of RSUs available in the market. In case of an accident, the accident message will be broadcast from the TMC to RSU, and RSUs will transmit the message to CVs. Based on the 300 m range of DSRC communication, RSUs can inform all the CVs inside that range. After defining the V2X environment in AIMSUN, the morning peak hour traffic demand (07:30–08:30 a.m.) was counted for a small area of the urban transport network in Brockton Massachusetts. A complete OD demand survey on the considered case study was conducted to determine the hourly traffic demand between every origin–destination pair. The case study is shown in Figure 1. Brockton is a city in Plymouth County, Massachusetts, and according to the United States Census Bureau, the city has a total area of 21.6 square miles (56 km²) (Brockton, MA, 2022).

Then, microscopic simulation was selected to simulate the current condition of the case study. In microscopic traffic simulations, vehicles are represented as separate agents, whose motion is governed by specific rules. Those agents may be in interaction, which also has an impact on their behavior. There are many well-established microscopic models for conventional vehicles, such as Gipps model (Gipps, 1981, 1976) (P.G. 1981), Wiedemann model (Fellendorf and Vortisch, 2010), Nagel–Schreckenberg model (Nagel and Schreckenberg, 1992) or Intelligent Driver model (Treiber et al., 2000). Gipps model was used to simulate conventional vehicles (non-CVs) in this research. Fifteen scenarios consisting of different MPR of connected cars and connected buses (0%–100%) were developed. Additionally, ten replications were considered for each scenario to ensure that the results are reliable. The number of replications has been determined by delay or density as a measure of effectiveness. A confidence interval of 95% and a tolerance error of 10% were considered to determine the exact number of required replications (Virginia department of transportation, 2013).

The attributes of CV car and CV bus (e.g. reaction time, standard gap, maximum acceleration, maximum deceleration and sensitivity factor) were determined based on the suggested values in (Huisman, 2015). Three types of messages (cooperative awareness message, MAP extended message and signal phase and timing extended message) were used to provide an appropriate connection between CV cars and CV buses. Based on these types of messages, CV cars and CV buses can be informed regarding the location of congestion and the location of signalized intersections. Therefore, they can promptly reroute, and that is why, the probability of this vehicle type being in the congestion areas will be mitigated.

4. Explanation of V2X software development kit

The V2X SDK was designed to allow a modeler to simulate autonomous and CVs. This Kit will be capable of creating a special network (the name is VANeT) to simulate different parts of autonomous and CVs in a fairly realistic environment. The relationship between the components of the V2X SDK is shown in Figure 2. Based on this figure, there are five components:

• Message type: Message type shows the types of various messages transmit/broadcast between vehicles (CV car and CV bus) and the RSUs. The V2X SDK implements some common standardized message types and some generic messages. When new types of messages are required for experimental applications, these may be added to the V2X SDK using the new message methods in the V2X application programming interface and passed between vehicles and RSUs in the same way as the existing message types.

• Channel: The communications channel is the simulated representation of the radio hardware and protocols that provide communication between vehicles. The default channel object in the V2X SDK is a simple range-based message-passing object with a stochastic probability of successful transmission. A channel represents a communications protocol used to pass information between vehicles and RSUs.

• On board unit (OBU): The OBU provides the receiver and transmitter in a vehicle, with the proportion of vehicles equipped with each type of OBU defined in the vehicle type. Each OBU is capable of using one or more channels to receive one or more message types on each channel.

• RSU: The RSU is the “I” component of the V2I communication network. An RSU has a physical location, connections to road network nodes, a set of channels it is able to use and a set of message types it is able to transmit and receive. It may also communicate with a TMC as one of a network of similar devices.

• TMC: The TMC is the integrator of the data from multiple RSUs and the initiator of coordinated signal control and traffic management actions. It communicates to the RSUs via a separate channel type, which may now be based on wired links or dedicated radio channels.

In AIMSUN Next, the physical and dynamic properties of CVs should be specified. After determining the parameters including the range of transmit/broadcast message(s), the location of RSU(s) on the road and the type of message(s) between TMS, RSUs and CVs, the modeler is capable of using the V2X SDK to simulate CV and AV vehicles.

CVs transmit and receive more information about their activity than conventional vehicles, and this information is also available to traffic control centers through intelligent transportation system infrastructure. This enables new forms of vehicle behavior through V2V communications, i.e. by platooning or by collaborative maneuver. After determining the aforementioned parameters, the modeler can use the V2X SDK to simulate CV and AV vehicles.

5. Explanation of the case study and detail of scenarios

A small area of the urban transport network in Brockton, Massachusetts, was considered as a case study. Residential, administrative and industrial land uses exist in the considered case study. Figure 3 shows the simulated area. All the centroids shown in Figure 3 have been numbered clockwise. In addition, the OD demand matrix for this case study is shown in Figure 4.

All the transit lines (bus lines), timing and phasing of signalized intersections and parking conditions were simulated the same as real world. The location of bus stops and the headway of transit lines (timetable) information were provided from Maps and Schedules (2022). The current traffic condition of the case study was modeled as the first scenario.

In terms of the number and location of RSUs, two RSUs with 300 m DSRC were defined. RSU locations are projected to provide the highest level of transmission and reception of messages between CVs. The location of RSUs and incident area is shown in Figure 5. We hypothesized an incident happens from 08:00 to 08:15 a.m. on the South-bound (SB) of commercial street. During a 15-min incident time, dynamic rerouting messages will transfer from the TMC to RSUs and from RSUs to CV cars and CV buses.

One of the applications of the CV is dynamic route guidance based on V2V, V2I and vehicle-to-smart terminal technologies. The CV guidance system facilitates dynamic guidance for road network flow using real-time traffic information. Traditional guidance systems have several shortcomings such as delays in incident detection, releasing guidance information and route planning due to inaccurate timing predictions. When an incident happens, the CV technology would inform the driver immediately via V2V and V2I technologies to reroute and avoid getting stuck in traffic congestion. The developed hybrid model in our study is able to assign non-CVs stochastically and assign CVs dynamically in a 1-h simulation interval. The rest of the inputs and assumptions used in the simulation are as follows:

(1) Overall simulation:

• Each scenario was run for 1 h of simulated time; ten runs with different random seeds were conducted, and the results were averaged. A 10-min warm-up period was used for all scenarios.

• The actualization of the trajectories is conducted every tenth of a second (10 Hz) for all the vehicles within the communication range.

• All the messages will transfer in 0.1 s inside the DSRC area.

• The simulated case study in AIMSUN software is 100% compliance to the real-word conditions in terms of speed and capacity of the roads, and the road gradient.

• The length of all non-CV and CV cars is 5 m (16.4 ft). Their width is 2.5 m (8.2 ft). The length of non-CV buses is 12 m (39.4 ft) and their width is 2.6 m (8.53 ft). The length of CV buses is 14m (45.9 ft) and their width is 2.6 m (8.53 ft).

(2) Gap-acceptance:

• The minimum lag between the front and rear parts of the conflicting vehicles is 1.0 s (Lioris et al., 2017).

The details of the scenarios are shown in Table (1). According to this table, traffic demand is quite the same in scenarios one, two and three. The only difference between scenarios one and two is the incident, and the only difference between scenarios two and three is the number of connected buses. The percentage of connected cars will increase from scenario 1 to 15. It means the percentage of non-CV (conventional) cars will mitigate, and the majority of substitute vehicles are able to reroute dynamically.

6. Analysis results

As it was mentioned before, we hypothesized the incident will happen from 08:00 to 08:15 a.m. in commercial link (SB). After the incident happens on this link, the majority of connected cars will divert to adjacent intersections. It is worth mentioning that all the CV cars in different approaches of “centre-commercial intersection” were traced. We found that almost 5% of connected cars prefer to use the incident link during 08:00–08:15 a.m. Figure 6 shows the delay time changes (sec/km) for all the vehicles. As shown in Figure 6, total delay time experienced a considerable increase from scenarios 1 to 2. The principal reason is a 15-min incident interval in scenario 2. A crash with a 15-min congestion time interval was considered on one of the cardinal links. This link was identified after investigating the speed, hourly volume and capacity of the link so that the critical congestion condition is obtained on the network. When CVs broadcast on the network (scenario 3), a slight improvement (5 sec/km) in total delay time is obtained. Figures 7 and 8 show the delay time changes for CV cars and CV buses.

The horizontal axes of Figure 6 show the percentage of CVs (third column of Table 1) in each scenario. As shown in Figure 6, delay time was enhanced from 113.7 in scenario 1 (without CV + no incident) to 145.4 (the peak value) in scenario 2 (without CV + incident), and then, it was mitigated to 140.4 in scenario 3 (with CV buses). The principal reason of delay time mitigation when the percentage is 0% (from scenarios 2 to 3) is “changing the input attributes of CV buses in the simulation” (consisting of max desired speed [km/hour], normal acceleration and deceleration [ ms2], overtaking maneuver [sec], reaction time [sec], average brake execution time [s], average max braking intensity and maximum give way time [sec]). These attributes were obtained for CV buses and considered in the microsimulation.

Delay time in urban networks is directly computed for each individual signalized intersection under speed estimation. Unsignalized intersection delay is computed by using an intersection approach delay formula for signalized intersections (USDOT, 2022). As shown in Figures 7 and 8, delay time was mitigated considerably for CV cars and CV buses. Dynamic rerouting of CV cars and CV buses helps them to distribute on the network efficiently, and platooning of CVs impressively improves the delay time on signalized and unsignalized network components.

According to Figures 6 to 8, when the percentage of connected cars increases, delay time (sec/km) mitigated due to the fact that CVs can dynamically reroute during the 15-min incident time interval. There are 11 signalized intersections in the considered case study, and they impose a considerable delay on CV and non-CV cars, but dynamically rerouting will assist CV cars to detract the total delay time. Delay time increases dramatically from the first to second scenarios because one of the principal links of the case study will not be served during a 15-min incident interval. From scenario 3, the connected buses will broadcast on the network. The third scenario would be the first point in reducing the total delay time. From scenario 4, the MPR of connected cars will increase, and they divert to upstream and downstream intersections. Nevertheless, a descending trend is obtained to scenario 15. Predominantly, two models are able to simulate CVs in AIMSUN. adaptive cruise control (ACC) is an advanced version of cruise control that is able to automatically maintain a certain set speed and also detect the speed of the vehicle in front and adapt to maintain a set distance. The second model is cooperative adaptive cruise control (CACC) that is a further development of ACC, which adds communications with multiple vehicles. Hence, CACC vehicles are able to send and receive speed information, enabling smoother and faster responses than ACC. These systems take over a part of the driving task, influencing the driving behavior of drivers and vehicles on the road. The ACC model was used in our research to simulate CV cars and CV buses. The ACC improves the process of speed and acceleration changes for CV cars and buses and will reduce the response time to traffic congestion in 15-min incident interval.

Delay time changes (hotmap) for non-CV cars, CV cars and CV buses are integrated as Figure 9. As shown in Figure 9, non-CV cars’ delay time is partially reduced, and CV cars and CV buses’ delay time is reduced considerably because of dynamic rerouting.

In terms of fuel consumption, Akcelic model (Akçelik, 1983) was considered. The fuel consumption model assumes that each vehicle is either idling or cruising at a constant speed or accelerating or decelerating. The state of each vehicle is determined, and the model then uses the appropriate formula to calculate the fuel consumed for this state. Equations (1) to (4) show the fuel consumption mathematical models:

(1) Fa=(C1+C2av)

(2) dFdt=k1(1+v32vm3)+k2v

(3) k1=(F1−F2)v1v2vm3180(2v2vm3−2v1vm3+v2v13−v1v23)

(4) k2=2F2v2vm3−2F1v1vm3+F2v2v13−F1v1v23360(2v2vm3−2v1vm3+v2v13−v1v23)

The following parameters are explained:

c₁ and c₂ = constants for accelerating vehicles in ml/s; a = vehicle acceleration;v = vehicle speed;v_m = the speed at which the fuel consumed per km is a minimum. Typically this is around 50 km/h;k₁ and k₂ = constants;

F₁ = the fuel consumption rate, in liters per 100 km, for vehicles traveling at a constant speed of 90 km/h; and

F₂ = the fuel consumption rate, in liters per 100 km, for vehicles traveling at a constant speed of 120 km/h.

Figure 10 shows the fuel consumption changes.

As shown in Figure 10, as the MPR of CVs increases, fuel consumption significantly decreases. The reason of decreasing the fuel consumption is dynamically rerouting of CVs. The proposed hybrid model in AIMSUN is able to figure out the shortest paths in terms of travel time. Therefore, the shortest paths are selected by CVs after crash. As shown in Figure 10, 5.2 liters enhancement in fuel consumption was seen from scenarios 1 to 2. A considerable percentage of non-CVs stop behind the incident area, and they exacerbate the fuel consumption changes from scenarios 1 to 2. The percentage of changes from scenarios 3 to 15 toward scenario 2 was calculated. The results showed that as the MPR of CVs in the network increases, the fuel consumption decreases by an average of 56%. Furthermore, the results highlighted that total speed of the network mitigated, the number of deceleration of non-CV cars increased and the number of CV cars’ and CV buses’ acceleration and decelerations increased due to the considerable number of signalized intersections and dynamic rerouting of CVs. Based on equations (1) to (4), as MPR of CVs increases, the total fuel consumption decreases for non-CV cars, CV cars and CV buses. It is worth mentioning that the software is not capable of prepare separate fuel consumption results for different vehicle types. Hence, an integrated fuel consumption changes chart was reported in Figure 10.

7. Future work

In future work, we intend to continue working on investigating the emission of pollutants. We intend to evaluate different environmental models and finally develop an environmental model based on carbon dioxide, particulate matter, nitrogen oxides and volatile organic compound emissions. Last but not least, it is better to evaluate the impression of CVs dynamic rerouting on the average travel times of other links. It means rerouting will impose additional travel time on non-CVs that are moving on the other links. Therefore, transportation equity should be investigated.

8. Conclusion

This paper attempted to quantify the potential positive effects of the gradual deployment of CVs through microscopic traffic simulation modeling. AIMSUN was used to model CVs, congestion areas and incidents associated with a small area of the urban transport network. The result of this research clearly demonstrated the effectiveness of CVs to improve delay time and fuel consumption. The MPR of CVs changes in the network is the most significant factor to improve total delay time and it can be explained by dynamically rerouting to alternate routes. This research attempted to use the V2X module in AIMSUN micro-simulation. Physical and behavioral features of CVs (car and bus) were defined. The number of RSUs, and their locations, were determined to the highest coverage of message transmission occurs in the considered case study. As soon as the incident happens, the message of congestion will be transmitted to RSUs, and then, RSUs will broadcast the message(s) to CVs. The travel time and location of CVs will be transmitted. Therefore, CVs can dynamically re-route on the network during the incident interval. Accordingly, new links will be used by CVs, and equilibrium is established in the network. The results showed that as MPR of CVs increases, the total delay time decreases by an average of 14% and the fuel consumption decreases by an average of 56%, respectively, from scenarios 3 to 15 compared to scenario 2.

Figures

Figure 1

Case study of this research

Figure 2

Architecture of V2X module in AIMSUN (AIMSUN Next 8.4.0, 2019)

Figure 3

Simulated case study in AIMSUN software

Figure 4

OD demand matrix

Figure 5

Location of RSUs and incident happening link

Figure 6

Total delay time (sec/km)

Figure 7

Delay time (sec/km) for connected cars (CV cars)

Figure 8

Delay time (sec/km) for connected bus (CV bus)

Figure 9

Integrated delay time (sec/km) chart for non-CV cars, CV cars and CV buses

Figure 10

Fuel consumption changes (liters)

Table 1

Detail of scenarios

Scenario	Incident	Total CVs		Total Non-CVs
		CV car (%)	CV bus (%)	Non-CV car (%)	Non-CV bus (%)
1	No	0	0	100	100
2	Yes	0	0	100	100
3		0	100	100	0
4		10	0	90	100
5		20	0	80	100
6		30	0	70	100
7		40	0	60	100
8		50	0	50	100
9		50	100	50	0
10		60	0	40	10
11		70	0	30	100
12		80	0	20	100
13		90	0	10	100
14		100	0	0	100
15		100	100	0	0