Fuel Consumption and Traffic Emissions Evaluation of Mixed Traffic Flow with Connected Automated Vehicles at Multiple Traffic Scenarios

Abstract

To analyze the impact of different proportions of connected automated vehicles (CAVs) on fuel consumption and traffic emissions, this paper studies fuel consumption and traffic emissions of mixed traffic flow with CAVs at different traffic scenarios. Firstly, the car-following modes and proportional relationship of vehicles in the mixed traffic flow are analyzed. On this basis, different car-following models are applied to capture the corresponding car-following modes. Then, Virginia Tech microscopic (VT-micro) model is adopted to calculate the instantaneous fuel consumption and traffic emissions. Finally, based on three typical traffic scenarios, a basic segment with bottleneck zone, ramp of the freeway, and signalized intersection, a simulation platform is built based on Python and SUMO to obtain vehicle trajectory data, and the fuel consumption and traffic emissions in different scenarios are obtained. The results show that (1) In different traffic scenarios, the application of CAVs can reduce fuel consumption and traffic emissions. The higher the penetration rate, the more significant the reduction in fuel consumption and traffic emissions. (2) In the three typical traffic scenarios, the advantages of CAVs are more evident in the signalized intersection. When the penetration rate of CAVs is 100%, the fuel consumption and traffic emissions reduction ratio is as high as 32%. It is noteworthy that the application of CAVs in urban transportation will significantly reduce fuel consumption and traffic emissions.

This work is licensed under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.

1. Introduction

With the development of the urban economy, the number of vehicles has increased rapidly. Traffic congestion has become more and more serious [1, 2], and also caused massive fuel consumption and traffic emissions [3]. In 2018, the United States consumed approximately 143 billion gallons of motor gasoline, with a daily average of 391 million gallons [4]. Thence, it is a top priority to reduce fuel consumption and emissions of traffic systems.

Autonomous driving technology is a comprehensive application of traffic monitoring, route navigation, and artificial intelligence technology. The development of connected automated transportation systems has shown a trend of integrating with autonomous driving technology. Therefore, CAVs [5, 6] will cause the transformation of intelligent transportation systems [7]. In the connected automated transportation system, the preceding vehicle can transmit position, speed, acceleration, and other information to the following vehicle in real-time through Vehicle-to-Vehicle (V2V) communication technology. At this time, CAVs can drive in cooperative adaptive cruise control (CACC) mode through V2V communication technology [8]. In addition, if V2V communication conditions are not met, CAVs will drive in adaptive cruise control (ACC) mode [9]. To sum up, the widespread application of CAVs will effectively reduce the response delay of vehicles to changes in surrounding traffic conditions and shorten the car-following time between vehicles. Therefore, CAVs can improve the inherent characteristics of traffic flow, and achieve a comprehensive breakthrough in alleviating traffic system congestion, energy-saving, and traffic emission reduction. Many scholars generally believe that the CAVs are expected to improve the quality of traffic flow from the microscopic traffic flow level [10, 11], and provide an effective way to bring down fuel consumption and traffic emissions.

In recent years, lots of studies have been focused on fuel consumption and emissions of the traffic system [12]. How to calculate the fuel consumption and emission of vehicles have always been a significant research topic. Therefore, some research focused on fuel consumption and emission models.

As the earliest generation of fuel consumption and emission factor model of motor vehicles, the Emission Factors (EMFAC) model [13] is developed by the California Air Resources Bureau. In addition, the European Environment Agency also funded the development of the computer program to calculate emissions from the road transport (COPERT) model [14]. Compared with EMFAC, the COPERT model has lower requirements on parameters, including 15 parameters such as fleet composition, average driving speed, average journey length, and fuel parameters, but its coverage is not as wide as EMFAC. Due to the design requirements for emissions, micro fuel consumption and emission models appeared. The Parametric Analytical Model of Vehicle Energy Consumption (PAMVEC) [15] calculates vehicle energy consumption based on total vehicle mass, parameter driving cycle description, and other attributes of the vehicle platform. The most intuitive description of the driving state of a motor vehicle is to establish a speed-acceleration matrix. The value of the matrix is the average fuel consumption level corresponding to the speed and acceleration. On this basis, Cernuschi et al. [16] divided the data into five types: high deceleration, low deceleration, uniform speed, high acceleration, and low acceleration. Meanwhile, the regression analysis method is used to fit best the emission speed curve of each type of acceleration mode, and the idle fuel consumption and emission are analyzed. Subsequently, under the premise of combining speed and acceleration with different power products, Ahn [17] proposed the VT-micro model, which determines the value of fuel consumption and different emission indicators based on the combination of different power products of speed and acceleration. With the development of CAVs, many researchers have begun to study the impact of CAVs on fuel consumption and traffic emissions. On this basis, Zegeye et al. provided a general framework to integrate macroscopic traffic flow models with microscopic fuel consumption and emission models, and extended the VT-micro model to the macroscopic traffic flow model to further shorten the simulation time and carry out fairly accurate estimates of the emissions and fuel consumption. Chandra and Camal [18] used simulation methods to calculate the fuel consumption and emissions of connected vehicle technology (CVT). The results indicated that the CVT reduces fuel consumption and emissions in different traffic conditions. Aiming at signalized intersections, Han et al. [19] used the trajectory optimization method, PTO-GFC, to study the fuel consumption and emissions of CAVs passing intersections under smooth trajectories. Research results showed that fuel consumption and emissions will be reduced to varying degrees under the proposed method. Yao et al. [6] studied the impact of mixed traffic flow at a single intersection on fuel consumption and emissions, and proposed a joint optimization framework for traffic signals and vehicle trajectories at a single intersection, including HDVs and CAVs. The results showed that this method could reduce fuel consumption and emissions. Yao and Li [20] applied a more detailed micro-fuel model (e.g., VT-micro model) in the study of CAVs trajectory optimization at an isolated signalized intersection with a single-lane road. Then, a joint objective of travel time, fuel consumption, and safety is implemented to improve traffic mobility, energy efficiency, and safety simultaneously. Aiming at the expressway ramp, Qin et al. [21] proposed a stability analysis method applied to mixed traffic flow, which simulates the mixed traffic flow of the expressway ramp segment to evaluate the impact of stability on fuel consumption and traffic emissions. Subsequently, the correlation between CACC vehicles improving traffic flow stability and reducing traffic emissions and fuel consumption was further studied. A simulation was applied to evaluate the changes in fuel consumption and emissions under traffic oscillation. The results showed that stability could qualitatively affect the reduction of fuel consumption and emissions, and provide new ideas for reducing fuel consumption and traffic emissions [22].

The above analysis indicated that the current research on fuel consumption and emission models is relatively mature. With the development of CAVs, related scholars have made useful explorations on the fuel consumption and emissions of the transportation system. However, to our best knowledge, there are some issues that need to be further discussed: (1) Some studies only focus on fuel consumption or emissions, and did not discuss the relationship between fuel consumption and traffic emissions under the same scenario. (2) Most researches on traffic scenarios are relatively single, and the impact mechanism of CAVs on fuel consumption and traffic emission under different traffic scenarios is not deeply discussed. (3) Most studies only focus on single-lane freeways. The scenarios do not involve lane-changing behavior, and the impact of vehicle lane-changing on fuel consumption and emission is not considered. Therefore, the impact of CAVs on the fuel consumption and traffic emissions of mixed traffic flows in different traffic scenarios remains to be revealed.

To address these issues, this paper is dedicated to evaluating the impact of the penetration rate of CAVs on the fuel consumption and traffic emission of mixed traffic flow under three classic traffic scenarios: basic road segment with bottleneck zone, a ramp of the freeway, and a signalized intersection. Firstly, the microscopic driving behavior and three car-following modes of mixed traffic flow are analyzed. Secondly, three car-following models were applied to describe these three car-following modes, and the lane-changing model LC2013 [23] was adopted for different traffic scenarios. Then, the VT-micro fuel consumption and emission model is introduced to calculate fuel consumption and traffic emissions in different traffic scenarios. Finally, based on a simulation experiment with SUMO, three traffic scenarios are simulated, and trajectory data are obtained. Based on the VT-micro model, fuel consumption and traffic emissions in different scenarios are calculated and analyzed. Thence, the main contributions of this work are as follows.

(1) The car-following behavior of vehicles in the mixed traffic flow is discussed, and different car-following models are used to describe the car-following characteristics in the mixed traffic flow.

(2) The impact of lane changing on fuel consumption and emission is considered, and lane-changing model LC2013 [23] is applied to different traffic scenarios.

(3) The influence of the penetration rate of CAVs on the fuel consumption and emission of mixed traffic flow under different traffic scenarios is studied.

The remainder of this paper is as follows. The micro-driving behavior, the car-following model, the car-following model, and the lane-changing model are analyzed in Section 2. Section 3 describes fuel consumption and emission models. Simulation analysis was conducted in Section 4, including simulation environment construction and simulation results analysis under different traffic scenarios. Finally, Section 5 summarizes the whole paper and proposes future work.

2. Micro Driving Behavior Analysis

2.1. Mixed Traffic Flow

2.1.1. Car-Following Modes

Considering that it will take a long time to upgrade and deploy CAVs, there will be a mixture of CAVs and HDVs on the road in the future [5, 24, 25]. As shown in Figure 1, there are four car-following modes: (i) HDV-HDV, (ii) HDV-CAV, (iii) CAV-HDV, and (iv) CAV-CAV.

[figure omitted; refer to PDF]

(1) Modes 1 and 2. As shown in Figure 1, the current vehicle of modes 1 and 2 is an HDV. When the front vehicle changes driving behavior (e.g., acceleration and deceleration), the driver of the current vehicle needs to perceive, identify, judge, and take action. Thence, car-following behavior is determined by the human driver.

(2) Mode 3. In this mode, the current and front vehicles are CAV and HDV, respectively. The current vehicle is equipped with the onboard sensing systems (e.g., laser, radar, camera), which can quickly and accurately capture the front vehicle’s driving behavior. This mode is called adaptive cruise control (ACC). The ACC system makes the current vehicle closely follow the front vehicle via controlling the accelerator and brake of the vehicle based on the advanced driving assistance system (ADAS).

(3) Mode 4. In this mode, the current and front vehicles are CAVs. The current vehicle can interact with the front vehicle via V2V communication. Therefore, the current vehicle can make synchronous driving behavior changes with the front vehicle. This mode is called cooperative adaptive cruise control (CACC). CACC system can shorten the distance between vehicles as much as possible on the premise of ensuring safety, to improve the traffic capacity of the transportation system.

2.1.2. Proportional Analysis

If it is assumed that the penetration rate of CAVs in the mixed traffic flow is $p$, then the proportion of HDVs is $1-p$. Thus, the proportions of car-following modes 1 and 2 are $1-p$. According to probability theory, the proportions of car-following modes 3 and 4 are $p1-p$ and $p^2$, respectively. To sum up, the proportions of different car-following modes are shown in Table 1.

Table 1

The ratio of the four modes.

Modes		Car-following models	Proportion
1	HDV-HDV	HDV	$1-p$
2	HDV-CAV
3	CAV-HDV	ACC	$p1-p$
4	CAV-CAV	CACC	$p^2$

2.2. Car-Following Model

In this paper, the intelligent driver model (IDM) [26], ACC, and CACC models [27, 28] are applied to describe the car-following behaviors of HDVs and CAVs. As shown in Figure 1, the IDM is utilized to simulate modes 1 and 2, and the ACC and CACC are applied to capture modes 3 and 4, respectively. The detailed description of the three car-following models is as follows.

2.2.1. Intelligent Driver Model

In the existing research, the models describing the driver’s car-following behavior mainly include stimulus-response model, safety distance model, social force model, and optimal speed model. As a kind of social force model, the IDM [26] can accurately describe the car-following behavior of drivers, and is widely used in the study of micro traffic flow. Therefore, the IDM model is adopted in this paper to describe the car-following behavior of HDVs. The model is defined by equation.$$\begin{array}{}\text{(1)}& a_{i,k+1}={\overline{a}}_{i}1-{\frac{v_{i,k}}{v_f}}^{\delta }-{\frac{s^{\ast }{v}_{i,k},\Delta v_{i,k}}{s_i}}^2,\end{array}$$where $a_{i,k+1}$ is the acceleration of the $i$-th vehicle at $k+1$ time step; ${\overline{a}}_i$ is the desired acceleration of the $i$-th vehicle; $v_f$ is the free flow velocity; $s^{\ast }$ is the expected distance between the $i-1$-th and -th vehicles; $v_{i,k}$ is the speed of the $i$-th vehicle at $k$ time step; $\mathrm{\Delta }{v}_{i,k}$ is the speed difference between the $i-1$-th and $i$-th vehicles at the $k$ time step; $\mathrm{\Delta }{v}_{i,k}=v_{i,k}-v_{i-1,k}$; $s_i$ is the actual distance between the $i-1$-th and $i$-th vehicles.

The expected distance $s^{\ast }$ is determined by the speed of the $i$-th vehicle and the speed difference of the preceding car, and the calculation equation is:$$\begin{array}{}\text{(2)}& s^{\ast }{v}_{i,k},\Delta v_{i,k}=s_0+T{v}_{i,k}+\frac{v_{i,k}\Delta v_{i,k}}{2\sqrt{{\overline{a}}_i{\overline{b}}_i}},\end{array}$$where $s_0$ is the minimum headway when stationary; $T$ is the safe time headway; ${\overline{b}}_i$ is the comfortable deceleration of the $i$-th vehicle. Referring to Treiber et al. [26], the parameters of IDM are: ${\overline{a}}_i=1$ m/s², ${\overline{b}}_i=2$ m/s², $T=1.5$s, $v_f=33.3$ m/s, and $s_0=2$ m.

2.2.2. Adaptive Cruise Control Model

The ACC system makes the current vehicle closely follow the front vehicle via controlling the accelerator and brake of the vehicle based on the advanced driving assistance system (ADAS). Generally, the ACC system will be equipped with sensing devices, such as laser, radar, camera, etc. These can measure the speed and position difference between the current vehicle and the front vehicle. The ACC system adjusts the current vehicle’s speed to keep the desired headway between it and the front vehicle. In particular, when there is no front vehicle, the ACC system will cruise at the designed speed.

The ACC car-following model [27, 28] developed and validated with field data is adopted in this paper. The integrated ACC car-following model is divided into four modes: (i) speed control, (ii) gap control, (iii) gap-closing control, and (iv) collision avoidance. The fourth mode (i.e., collision avoidance) is developed by TransAID [23], which tries to avoid rear-end collisions between adjacent vehicles. The details of the four control modes are discussed below.

(1) Speed Control. The speed control mode aims to keep the preset speed of the vehicle. When the vehicle has no preceding vehicle within the coverage of the sensor or the distance from the preceding vehicle is more significant than 120 m, the speed control mode is activated [28]. The speed control mode calculates the acceleration of the $i$-th vehicle at $k$ time step and is given by the following equation.$$\begin{array}{}\text{(3)}& a_{i,k+1}={\beta }_1{v}_d-v_{i,k},\end{array}$$where $v_d$ is the ideal velocity; $v_{i,k}$ is the velocity of the $i$ th vehicle at $k$ time step; and ${\beta }_1$ is the control gain, which determines the speed deviation rate of acceleration. Xiao et al. [28] suggested that the value is 0.3-0.4 $s^{-1}$. In this paper, the value of 0.4 $s^{-1}$ was selected.

(2) Gap Control. The gap control mode keeps a constant time gap between the vehicle equipped with ACC and the vehicle ahead. This mode is activated when the gap and speed deviations (vehicle in front) are, respectively, less than 0.2 m and 0.1 m/s. The acceleration of the $i$-th vehicle at time step $k+1$ is described by the second-order function based on the velocity and gap error with the leading vehicle, which is defined in the following equation.$$\begin{array}{}\text{(4)}& a_{i,k+1}={\beta }_2{e}_{i,k}+{\beta }_3{v}_{i-1,k}-v_{i,k},\end{array}$$where $e_{i,k}$ is the gap error of the $i$ th vehicle at $k$ time step; ${\beta }_2$ and ${\beta }_3$ are the control gain for position and velocity deviations. Referring to Xiao et al. [28], the calibration values of control parameters in equation (4) are ${\beta }_2=0.23\hspace{0.17em}{s}^{-1}$ and ${\beta }_3=0.07\hspace{0.17em}{s}^{-1}$. Furthermore, the gap error $e_{i,k}$ is given by$$\begin{array}{}\text{(5)}& e_{i,k}=x_{i-1,k}-x_{i,k}-{\tau }_a{v}_{i,k},\end{array}$$where $x_{i-1,k}$ and $x_{i,k}$ are the positions of the $i-1$-th and the $i$-th vehicles at time step $k$, and the ${\tau }_d$ is the desired time gap for the ACC system; its value range is generally 1.1 to 2.2 s. In this study, the value of ${\tau }_a$ is set to 1.3 s.

(3) Closed Gap Control. The gap closure controller realizes the stable transition from speed control mode to gap control mode. This mode is activated when the distance to the vehicle in front is less than 100 m. According to Xiao et al. [28], the gap closure control mode can be presented by setting parameters of equation (4) as ${\beta }_2=0.04\hspace{0.17em}{s}^{-1}$ and ${\beta }_3=0.8\hspace{0.17em}{s}^{-1}$. Specially, if the distance is between 100 m and 120 m, ACC-equipped vehicles will retain the fore control mode to supply hysteresis in the control loop and stable transition between the two modes.

(4) Collision Avoidance Control. The collision avoidance controller is developed by the TransAID team to prevent vehicle rear-end collision under safe conditions. When the distance to the preceding vehicle is less than 100 m, and the position gap deviation is negative, this mode is activated. Similar to gap-closing control, the collision avoidance control mode can be presented by setting parameters of equation (4) as ${\beta }_2=0.8\hspace{0.17em}{s}^{-1}$ and ${\beta }_3=0.23\hspace{0.17em}{s}^{-1}$. In simulation, these parameters can ensure the safe operation of ACC vehicles, even in critical events [23].

2.2.3. Cooperative Adaptive Cruise Model

Compared with the ACC system, the CACC system can communicate with each other through V2V communication. Therefore, the CACC system can further improve traffic safety and traffic efficiency. The integrated CACC car-following model is verified by the actual data based on the work of Milanés et al. [29] and Xiao et al. [28, 30]. Similar to the ACC system, the developed control algorithm in the CACC system is divided into four modes: (i) speed control, (ii) gap control, (iii) closed gap control, and (iv) collision avoidance control.

(1) Speed Control. The speed control mode aims to keep the desired speed predefined by the driver. This mode is activated when there are no preceding vehicles in the sensor coverage area or the time gap is greater than 2 seconds. The acceleration of the $i$-th vehicle at time step $k+1$ can be obtained by equation.$$\begin{array}{}\text{(6)}& a_{i,k+1}={\beta }_4{v}_d-v_{i,k},\end{array}$$where ${\beta }_4$ is the speed control gain. Referring to Milanés and Shladover [27], the value of ${\beta }_4$ is set as $0.4\hspace{0.17em}{s}^{-1}$.

(2) Gap Control. The gap control mode keeps a constant time gap between the vehicle equipped with CACC and the vehicle ahead. This mode is activated when the gap and velocity deviations (vehicle in front) are, respectively, less than 0.2 m and 0.1 m/s. The gap control of the CACC model is based on the first-order transfer function of velocity and gap error, which can be defined in the following equation.$$\begin{array}{}\text{(7)}& v_{i,k}=v_{i,k}+{\beta }_5{e}_{i,k}+{\beta }_6{\dot{e}}_{i,k},\end{array}$$where ${\dot{e}}_{i,k}$ is the first derivative of the gap error $e_{i,k}$. The values of ${\beta }_5$ and ${\beta }_6$ are set to ${\beta }_5=0.45\hspace{0.17em}{s}^{-1}$ and ${\beta }_6=0.0125\hspace{0.17em}{s}^{-1}$, respectively [31]. The first derivative of the gap error is given by$$\begin{array}{}\text{(8)}& {\dot{e}}_{i,k}=v_{i-1,k}-v_{i,k}-{\tau }_d{a}_{i,k},\end{array}$$where $a_{i,t}$ is the acceleration at time $t$, and ${\tau }_d$ is the required time gap defined by the CACC controller. The value is 0.6 to 2.2 s, and the value of ${\tau }_d$ in this study is set to 1.0 s.

(3) Closed Gap Control. The gap closure control mode realizes a stable transition from the speed control mode to the gap control mode. This mode is activated when the time distance to the front of the vehicle is less than 1.5 s. The control formula is consistent with equation (7), but the value of the gain coefficient parameter is inconsistent. Referring to Reference [23], the gain coefficient parameter values are ${\beta }_5=0.005\hspace{0.17em}{s}^{-1}$ and ${\beta }_6=0.05\hspace{0.17em}{s}^{-1}$.

(4) Collision Avoidance Control. The collision avoidance mode prevents vehicle rear-end collision under safe conditions. When the time gap is less than 1.5 s, and the position gap deviation is negative, this mode is activated. The controller’s logic is consistent with the gap control and could be expressed by changing the gain coefficients, which are ${\beta }_5=0.45\hspace{0.17em}{s}^{-1}$ and ${\beta }_6=0.05\hspace{0.17em}{s}^{-1}$ [23].

In summary, both ACC and CACC car-following models contain four control modes, and the parameter values of different control modes can be summarized as shown in Figure 2.

[figure omitted; refer to PDF]

The result shows that CAVs can maintain a more stable driving speed and minor headway in the road segments. Meanwhile, CACC vehicles form a platoon during operation. Compared with traditional HDVs, the acceleration and deceleration processes are significantly reduced, thus reducing the fuel consumption and emission of traffic flow. However, in this scenario, when vehicles run near the ramp, an HDV or a CAV will be disturbed by the vehicle in the ramp, resulting in different degrees of acceleration and deceleration behavior. Therefore, in the ramp of the freeway, when the penetration rate of CAVs is 100%, compared with 0%, the decline in fuel consumption, CO, HC, and NOx are relatively low, all lower than 12%.

4.2.3. Signalized Intersection

In urban transportation, the signalized intersection is an essential node for vehicle collection, turning, and evacuation. The setting of signalized intersection scenario is of great significance to the study of the characteristics of traffic flow. The scenario here is a two-way three-lane intersection, including three turns: left turn, straight, and right turn. The traffic flow rate of each lane is 220 veh/h. Based on the simulation experiments, the fuel consumption and traffic emissions under different CAVs conditions in this scenario are obtained, as shown in Table 7.

Table 7

Fuel consumption and emissions under different penetration rates of CAVs (Scenario 3).

Permeability (%)	FC (L)	CO (g)	HC (g)	NOx (g)
0	29.74	167.38	27.35	22.47
10	28.55	161.80	26.21	21.66
20	27.32	156.24	25.05	20.84
30	26.35	152.22	24.15	20.19
40	25.27	147.76	23.15	19.48
50	24.00	142.40	22.01	18.64
60	22.66	137.03	20.84	17.74
70	21.51	132.04	19.90	16.98
80	20.12	125.67	18.91	16.03
90	18.96	120.11	18.21	15.22
100	17.58	112.36	17.49	14.24

Table 7 shows that when the penetration rate of CAVs is 0%, the fuel consumption, CO, HC, and NOx are 29.74 L, 167.38 g, 27.35 g, and 22.47 g, respectively; when the permeability is 100%, the fuel consumption and traffic emissions are 17.58 L, 112.36 g, 17.49 g, and 14.24 g, respectively. In scenario 3, as the penetration rate of CAVs increases, fuel consumption and traffic emissions are all showing a downward trend. The changes in fuel consumption and traffic emissions under different penetration rates are shown in Figure 9. Compared with fuel consumption, HC, and NOx, CO changes more significantly, and the impact of CAVs on its emission is noticeable. In addition, compared with scenarios 1 and 2, the fuel consumption and traffic emissions in scenario 3 show a linear decrease as the penetration rate of CAVs increases. This means that the use of CAVs can significantly improve fuel consumption and traffic emissions in scenario 3.

[figures omitted; refer to PDF]

Furthermore, the reduction percentages of fuel consumption and traffic emissions under different penetration rates of CAVs is obtained, as shown in Table 8. Meanwhile, Figure 10 provides a more intuitive understanding of the impact of the penetration rate of CAVs on fuel consumption and traffic emissions. It can be seen from Figure 10 that as the penetration rate of CAVs increases, various fuel consumption and traffic emission indicators show a downward trend, and the decline is relatively significant. In scenario 3, the reduction of CO among the four indicators is slightly lower. When the penetration rate of CAVs is 100%, the percentage of CO reduction will reach 32.87%, while the declining percentage of fuel consumption, HC, and NOx reach 40.89%, 36.06%, and 36.64%, respectively.

Table 8

Percentage reduction in fuel consumption and traffic emissions (Scenario 3).

Penetration rate (%)	FC (%)	CO (%)	HC (%)	NOx (%)
0	NA	NA	NA	NA
10	4.01	3.33	4.16	3.59
20	8.14	6.65	8.41	7.27%
30	11.39	9.05	11.71	10.13
40	15.01	11.72	15.36	13.29
50	19.29	14.92	19.55	17.04
60	23.80	18.13	23.80	21.04
70	27.66	21.11	27.23	24.44
80	32.34	24.92	30.88	28.66
90	−36.25	−28.24	−33.43	−32.28
100	−40.89	−32.87	−36.06	−36.64

[figure omitted; refer to PDF]

In scenario 3, the road is relatively more standardized, and there are traffic lights to restrict the traffic flow. Therefore, in this scenario, the advantages of CAVs are more prominent. The four evaluation indicators have a significant decline, reaching more than 32% when the penetration rate of CAVs is 100%.

5. Conclusion and Policy Implication

This paper analyzes the fuel consumption and traffic emissions of mixed traffic flow with CAVs in different traffic scenarios. According to the results of the simulation experiment, the conclusions can be obtained as follows:

(1) In different traffic scenarios, CAVs play an effective role in reducing fuel consumption and emissions. With the increase of the penetration rate of CAVs, the declining proportion of fuel consumption and traffic emission increases. This suggests that the widespread application of CAVs can reduce fuel consumption and traffic emission, save resources, and alleviate environmental pollution to a certain extent.

(2) There are certain differences in the impact of CAVs on fuel consumption and traffic emission in different traffic scenarios. In the signalized intersection, the percentage of fuel consumption and traffic emissions reduction both reached more than 32%. However, in the ramp and bottleneck zone of the freeway, when the penetration rate of CAVs is 100%, both fuel consumption and traffic emissions decreased by less than 20%.

(3) Compared with the ramp and bottleneck zone of the freeway, the reduction in fuel consumption and traffic emissions of the signalized intersection is more significant, up to 32% or more. This means that the application of CAVs has more significant energy-saving and emission-reduction effects on urban transportation.

This paper mainly analyzes the mixed traffic flow’s fuel consumption and emissions with CAVs and HDVs based on the micro-driving behavior. Thus, the optimization and coordinated control of the CAVs are not considered in this study. Therefore, the collaborative control and optimization of connected automated vehicles can be considered further to reduce fuel consumption and traffic emissions in future work [37, 38]. Moreover, the trajectory of the CAVs can also be optimized in real-time to achieve the effect of alleviating traffic congestion, fuel consumption, and traffic emissions.

Acknowledgments

The paper received research funding support from the National Natural Science Foundation of China (52002339), the Sichuan Science and Technology Program (2021YJ0535, 2020YFH0026), the Fundamental Research Funds for the Central Universities (2682021CX058), the Guangxi Science and Technology Program (2021AA01007AA), and the Innovation Center Project of Chengdu Jiao Da Big Data Technology Co., Ltd. (JDSKCXZX202003).

References

[1] L. Yang, K. W. Chau, W. Y. Szeto, X. Cui, X. Wang, "Accessibility to transit, by transit, and property prices: spatially varying relationships," Transportation Research Part D: Transport and Environment, vol. 85,DOI: 10.1016/j.trd.2020.102387, 2020.

[2] L. Yang, X. Chu, Z. Gou, H. Yang, Y. Lu, W. Huang, "Accessibility and proximity effects of bus rapid transit on housing prices: heterogeneity across price quantiles and space," Journal of Transport Geography, vol. 88,DOI: 10.1016/j.jtrangeo.2020.102850, 2020.

[3] Y. Jiang, B. Zhao, M. Liu, Z. Yao, "A two-level model for traffic signal timing and trajectories planning of multiple CAVs in a random environment," Journal of Advanced Transportation, vol. 2021,DOI: 10.1155/2021/9945398, 2021.

[4] Energy Information Administration, "Use of gasoline - energy explained, your guide to understanding energy - energy information administration," 2018. https://www.eia.gov/energyexplained/

[5] Z. Yao, T. Xu, Y. Jiang, R. Hu, "Linear stability analysis of heterogeneous traffic flow considering degradations of connected automated vehicles and reaction time," Physica A: Statistical Mechanics and Its Applications, vol. 561,DOI: 10.1016/j.physa.2020.125218, 2021.

[6] Z. Yao, H. Jiang, Y. Cheng, Y. Jiang, B. Ran, "Integrated schedule and trajectory optimization for connected automated vehicles in a conflict zone," IEEE Transactions on Intelligent Transportation Systems,DOI: 10.1109/tits.2020.3027731, 2020.

[7] H. S. Mahmassani, "50th anniversary invited article-autonomous vehicles and connected vehicle systems: flow and operations considerations," Transportation Science, vol. 50 no. 4, pp. 1140-1162, DOI: 10.1287/trsc.2016.0712, 2016.

[8] S. E. Shladover, C. Nowakowski, X. Y. Lu, R. Ferlis, "Cooperative adaptive cruise control," Transportation Research Record: Journal of the Transportation Research Board, vol. 2489 no. 1, pp. 145-152, DOI: 10.3141/2489-17, 2015.

[9] M. Wang, S. P. Hoogendoorn, W. Daamen, B. van Arem, B. Shyrokau, R. Happee, "Delay-compensating strategy to enhance string stability of adaptive cruise controlled vehicles," Transportation Business: Transport Dynamics, vol. 6 no. 3, pp. 211-229, DOI: 10.1080/21680566.2016.1266973, 2018.

[10] J. Rios-Torres, A. A. Malikopoulos, "A survey on the coordination of connected and automated vehicles at intersections and merging at highway on-ramps," IEEE Transactions on Intelligent Transportation Systems, vol. 18,DOI: 10.1109/tits.2016.2600504, 2017.

[11] R. Scarinci, B. Heydecker, "Control concepts for facilitating motorway on-ramp merging using intelligent vehicles," Transport Reviews, vol. 34 no. 6, pp. 775-797, DOI: 10.1080/01441647.2014.983210, 2014.

[12] Z. Yao, Y. Wang, B. Liu, B. Zhao, Y. Jiang, "Fuel consumption and transportation emissions evaluation of mixed traffic flow with connected automated vehicles and human-driven vehicles on expressway," Energy, vol. 230,DOI: 10.1016/j.energy.2021.120766, 2021.

[13] "California Air Resources Board," 2002. WWW Document

[14] L. Ntziachristos, Z. Samaras, COPERT III, Computer Programme to Calculate Emissions from Road Transport, 2000.

[15] A. Simpson, Parametric Modelling of Energy Consumption in Road Vehicles, 2005.

[16] S. Cernuschi, M. Giugliano, A. Cemin, I. Giovannini, "Modal analysis of vehicle emission factors," The Science of the Total Environment, vol. 169,DOI: 10.1016/0048-9697(95)04646-i, 1995.

[17] K. Ahn, Microscopic Fuel Consumption and Emission Modeling (Thesis), 1998.

[18] S. Chandra, F. Camal, "A simulation-based evaluation of connected vehicle technology for emissions and fuel consumption," Procedia Engineering, vol. 145, pp. 296-303, DOI: 10.1016/j.proeng.2016.04.077, 2016.

[19] X. Han, R. Ma, H. M. Zhang, "Energy-aware trajectory optimization of CAV platoons through a signalized intersection," Transportation Research Part C: Emerging Technologies, vol. 118,DOI: 10.1016/j.trc.2020.102652, 2020.

[20] H. Yao, X. Li, "Decentralized control of connected automated vehicle trajectories in mixed traffic at an isolated signalized intersection," Transportation Research Part C: Emerging Technologies, vol. 121,DOI: 10.1016/j.trc.2020.102846, 2020.

[21] Y. Qin, H. Wang, B. Ran, "Stability analysis of connected and automated vehicles to reduce fuel consumption and emissions," Journal of Transportation Engineering, Part A: Systems, vol. 144 no. 11,DOI: 10.1061/jtepbs.0000196, 2018.

[22] Y. Qin, H. Wang, B. Ran, "Impacts of cooperative adaptive cruise control platoons on emissions under traffic oscillation," Journal of Intelligent Transportation Systems, vol. 25 no. 4, pp. 376-383, DOI: 10.1080/15472450.2019.1702534, 2021.

[23] E. Mintsis, D. Koutras, K. Porfyri, E. Mitsakis, S. Boerma, Modelling, Simulation and Assessment of Vehicle Automations and Automated Vehicles’ Driver Behaviour in Mixed Traffic, 2018.

[24] Z. Yao, R. Hu, Y. Wang, Y. Jiang, B. Ran, Y. Chen, "Stability analysis and the fundamental diagram for mixed connected automated and human-driven vehicles," Physica A: Statistical Mechanics and Its Applications, vol. 533,DOI: 10.1016/j.physa.2019.121931, 2019.

[25] Z. Yao, R. Hu, Y. Jiang, T. Xu, "Stability and safety evaluation of mixed traffic flow with connected automated vehicles on expressways," Journal of Safety Research, vol. 75, pp. 262-274, DOI: 10.1016/j.jsr.2020.09.012, 2020.

[26] M. Treiber, A. Hennecke, D. Helbing, "Congested traffic states in empirical observations and microscopic simulations," Physical Review, vol. 62 no. 2, pp. 1805-1824, DOI: 10.1103/physreve.62.1805, 2000.

[27] V. Milanés, S. E. Shladover, "Modeling cooperative and autonomous adaptive cruise control dynamic responses using experimental data," Transportation Research Part C: Emerging Technologies, vol. 48, pp. 285-300, DOI: 10.1016/j.trc.2014.09.001, 2014.

[28] L. Xiao, M. Wang, B. van Arem, "Realistic car-following models for microscopic simulation of adaptive and cooperative adaptive cruise control vehicles," Transportation Research Record: Journal of the Transportation Research Board, vol. 2623 no. 1,DOI: 10.3141/2623-01, 2017.

[29] V. Milanés, S. E. Shladover, J. Spring, C. Nowakowski, H. Kawazoe, M. Nakamura, "Cooperative adaptive cruise control in real traffic situations," IEEE Transactions on Intelligent Transportation Systems, vol. 15 no. 1, pp. 296-305, DOI: 10.1109/tits.2013.2278494, 2014.

[30] L. Xiao, M. Wang, W. Schakel, B. van Arem, "Unravelling effects of Cooperative Adaptive Cruise Control deactivation on traffic flow characteristics at merging bottlenecks," Transportation Research Part C: Emerging Technologies, vol. 96, pp. 380-397, DOI: 10.1016/j.trc.2018.10.008, 2018.

[31] H. Liu, X. Kan, S. E. Shladover, X.-Y. Lu, R. E. Ferlis, "Impact of cooperative adaptive cruise control on multilane freeway merge capacity," Journal of Intelligent Transportation Systems, vol. 22 no. 3, pp. 263-275, DOI: 10.1080/15472450.2018.1438275, 2018.

[32] K. Ahn, H. Rakha, A. Trani, M. Van Aerde, "Estimating vehicle fuel consumption and emissions based on instantaneous speed and acceleration levels," Journal of Transportation Engineering, vol. 128, pp. 182-190, DOI: 10.1061/(asce)0733-947x(2002)128:2(182), 2002.

[33] R. Akcelik, "Efficiency and drag in the power-based model of fuel consumption," Transportation Research Part B: Methodological, vol. 23 no. 5, pp. 376-385, DOI: 10.1016/0191-2615(89)90014-3, 1989.

[34] J. N. Hooker, "Optimal driving for single-vehicle fuel economy," Transportation Research Part Policy Pract, vol. 22, pp. 183-201, DOI: 10.1016/0191-2607(88)90036-2, 1988.

[35] H. Rakha, K. Ahn, "Closure to “estimating vehicle fuel consumption and emissions based on instantaneous speed and acceleration levels” by kyoung Ahn, hesham rakha, antonio trani, and michel van aerde," Journal of Transportation Engineering, vol. 129 no. 5, pp. 579-581, DOI: 10.1061/(asce)0733-947x(2003)129:5(579), 2003.

[36] T. Q. Tang, H.-J. Huang, H. Y. Shang, "Influences of the driver’s bounded rationality on micro driving behavior, fuel consumption and emissions," Transportation Research Part D: Transport and Environment, vol. 41, pp. 423-432, DOI: 10.1016/j.trd.2015.10.016, 2015.

[37] Z. Yao, B. Zhao, T. Yuan, H. Jiang, Y. Jiang, "Reducing gasoline consumption in mixed connected automated vehicles environment: a joint optimization framework for traffic signals and vehicle trajectory," Journal of Cleaner Production, vol. 265,DOI: 10.1016/j.jclepro.2020.121836, 2020.

[38] S. K. Zegeye, B. De Schutter, J. Hellendoorn, E. A. Breunesse, A. Hegyi, "Integrated macroscopic traffic flow, emission, and fuel consumption model for control purposes," Transportation Research Part C: Emerging Technologies, vol. 31, pp. 158-171, DOI: 10.1016/j.trc.2013.01.002, 2013.