0
Select Articles

Evaluating Connected Vehicles and Their Applications PUBLIC ACCESS

[+] Author Notes
Mohd Azrin Mohd Zulkefli, Pratik Mukherjee, Yunli Shao, Zongxuan Sun

Department of Mechanical Engineering University of Minnesota

Mohd Azrin Mohd Zulkefli received the B.S. and M.S. degrees in Mechanical Engineering from the University of Minnesota, Twin Cities, in 2006 and 2009, respectively. He is currently pursuing the Ph.D. degree with the Automotive Propulsion Control Group in Mechanical Engineering at the University of Minnesota, Twin Cities. His research interests are optimization and evaluation of hybrid vehicles in a connected vehicle setting.

Pratik Mukherjee received the B.S. and M.S. degrees in Mechanical Engineering from the University of Minnesota, Twin Cities in 2014 and 2016, respectively. He is currently pursuing the Ph.D. degree with the Coordination at Scale Lab in Electrical Engineering at the Virginia Polytechnic Institute, Blacksburg. His current research interest is multi-agent topological control for autonomous robots.

Yunli Shao received the B.S. degree in Mechanical Engineering from Shanghai Jiao Tong University, Shanghai, China, in 2013, and the M.S. degree in Mechanical Engineering from the University of Michigan, Ann Arbor, in 2015. He is currently working toward the Ph.D. degree with the Automotive Propulsion Control Lab in Mechanical Engineering at the University of Minnesota, Twin Cities. His research interests are control, optimization and evaluation of connected vehicles.

Zongxuan Sun received the B.S. degree in automatic control from Southeast University, Nanjing, China, in 1995, and the M.S. and Ph.D. degrees in mechanical engineering from the University of Illinois at Urbana–Champaign, Champaign, IL, USA, in 1998 and 2000, respectively. He is currently Professor of Mechanical Engineering at the University of Minnesota, Minneapolis, MN, USA, and the Co-Deputy Director of the NSF Engineering Research Center of Compact and Efficient Fluid Power. He was a Staff Researcher (2006–2007) and a Senior Researcher (2000–2006) at General Motors Research and Development Center, Warren, MI, USA. He has published more than 110 refereed technical papers and received 19 U.S. patents. His research interests include controls and mechatronics with applications to automotive propulsion systems.

Mechanical Engineering 138(12), S12-S17 (Dec 01, 2016) (5 pages) Paper No: ME-16-DEC3; doi: 10.1115/1.2016-Dec-3

This article presents evaluation results of connected vehicles and their applications. Vehicle-to-vehicle communication (V2V) and vehicle-to-infrastructure communication (V2I) can enable a new paradigm of vehicle applications. The connected vehicle applications could significantly improve vehicle safety, mobility, energy savings, and productivity by utilizing real-time vehicle and traffic information. In the foreseeable future, connected vehicles need to operate alongside unconnected vehicles. This makes the evaluation of connected vehicles and their applications challenging. The hardware-in-the-loop (HIL) testbed can be used as a tool to evaluate the connected vehicle applications in a safe, efficient, and economic fashion. The HIL testbed integrates a traffic simulation network with a powertrain research platform in real time. Any target vehicle in the traffic network can be selected so that the powertrain research platform will be operated as if it is propelling the target vehicle. The HIL testbed can also be connected to a living laboratory where actual on-road vehicles can interact with the powertrain research platform.

Vehicle to vehicle communication (V2V) and vehicle to infrastructure communication (V2I) can enable a new paradigm of vehicle applications. This paradigm shift will have profound impact on passenger and commercial vehicles, including both on-road and offroad vehicles. The connected vehicle applications could significantly improve vehicle safety, mobility, energy savings, and productivity by utilizing real-time vehicle and traffic information [1-2]. To realize this vision, three basic elements are required. First, communication devices and sensors need to be installed on both vehicles and the infrastructure. Second, new algorithms need to be developed to coordinate vehicle operation through V2V and V2I communication. Third, systematic and efficient testing methods are required to evaluate various connected vehicle applications (CVAs) to ensure their safety and effectiveness. This article focuses on the third element: evaluation of connected vehicles (CVs).

The average life span of automobiles in the United States is 15-20 years. This fact indicates that it will take at least 20 years for all vehicles on the road to be equipped with communication devices. Before then, on-road transportation will operate in the so-called “multimodal” mode. Vehicles with or without on-board communication devices need to operate together. Road infrastructure such as roadside equipment (RSE) will also be deployed gradually. This will be coupled with various vehicle automation levels, such as no automation (human operated), partial automation, or fully automated vehicles. The multimodal operation makes the evaluation of connected vehicles challenging, especially considering the safety factors for testing vehicles in real traffic. Given the above challenges, it is crucial to develop systematic, safe, and efficient testing methods for connected vehicles and their applications.

Currently the performance of a vehicle's fuel economy and emissions in real traffic is measured through either simulation or by instrumenting the vehicle with on-board instruments. There are limitations and challenges for either of the existing approaches. First, a simulation-based approach replaces the engine with steady-state fuel-use and emission maps and therefore may not be able to capture the transient behaviors. Secondly, instrumenting vehicles is time consuming and expensive since it requires major modifications of the vehicle. In addition, equipping large precision measurement devices on small passenger vehicles is challenging for testing purposes. One idea to solve this problem is to develop hardware-in-the-loop (HIL) testbeds for evaluating connected vehicles. The HIL system utilizes a real engine for direct fuel and emission measurements. Furthermore, different vehicles can be tested quickly and flexibly by changing the engine and the load settings on the dynamometer. The HIL testbed can also accommodate large precision measurement devices since it is built in a laboratory setting. Testing connected vehicle applications in simulated but realistic traffic is more economical without having to instrument multiple vehicles. It is also safer and bypasses the legal issues that would otherwise hamper the evaluation of connected vehicle applications in real traffic.

In this section, we will present a HIL testbed for evaluating connected vehicles at the University of Minnesota (UMN). The main objective of this testbed is to integrate a powertrain research platform [3] with a real-time traffic simulator (VISSIM) so that we can operate the powertrain research platform as if it is propelling any target vehicle selected from the simulated traffic network. This allows systematic evaluation of connected vehicle mobility and energy savings precisely, safely and efficiently. The architecture of the HIL testbed is shown in Figure 1.

FIGURE 1 HIL Testbed Architecture.

Grahic Jump LocationFIGURE 1 HIL Testbed Architecture.

VISSIM is a microscopic traffic simulator that allows the user to access traffic simulation states, such as vehicle speed, road conditions, and signal phase and timing, at every simulation time-step. Traffic networks can be set up in VISSIM to simulate real-world traffic scenarios. Field data can be used to calibrate the VISSIM simulation. A signal control cabinet can be connected to VISSIM to control the signal phase and timing in the traffic network. The connected vehicle controller is used to evaluate different vehicle control strategies such as cooperative adaptive cruise control (CACC). The powertrain research platform is used to emulate the powertrain operation of any target vehicle selected in VISSIM. Once the target vehicle speed and acceleration are received in real-time, the vehicle power demand is calculated and the powertrain operation is determined. The laboratory fuel and emissions measurement instruments are used to measure the fuel consumption and emissions in real-time.

The HIL testbed can be further enhanced by a living laboratory. The living laboratory enables the HIL testbed to represent a vehicle that follows the on-road test-vehicle to evaluate the performance of CVAs. To enable real-time communication, a road segment will be equipped with the roadside equipment and a small number of test-vehicles will be instrumented with onboard units (OBU). An OBU collects the test-vehicle data (e.g., location, speed, acceleration, pedal position, etc.) and generates a basic safety message (BSM) to broadcast to the RSE and other test-vehicles. RSE broadcasts traffic control signal phase and timing (SPaT) and MAP data to test vehicles, and forwards BSM and SPaT to the HIL testbed. With the equipped road and test vehicles, vehicle data can therefore be transmitted to the HIL testbed in real time to coordinate the on-road test-vehicles and the HIL testbed.

In this section, we show an example of using the HIL testbed to evaluate the performance of an emerging connected vehicle application: cooperative adaptive cruise control (CACC) [4]. CACC is intended to form a platoon of vehicles where each vehicle (except the lead vehicle) can automatically follow the preceding vehicle with safe separation distance by using radar or other sensors to measure the distance to the preceding vehicle and a communication device to communicate with other vehicles or the infrastructure. To evaluate the CACC application, as shown in Figure 1, the traffic simulator, connected vehicle controller (the CACC controller), and the powertrain research platform are connected in real time with the help of the middleware. Assuming that perfect communication is available between vehicles in a traffic network, the CACC algorithm is able to obtain preceding vehicles’ attributes like vehicle speed to derive the velocity of the CACC vehicle for the next time step.

A platoon of 15 vehicles is simulated in the VISSIM traffic simulator where the 15th vehicle is controlled by the external CACC controller and the other 14 preceding vehicles are controlled by the internal driver model of VISSIM. The platoon of vehicles is simulated on a traffic network that emulates a local highway as shown in Figure 2. During every time step of the simulation, the vehicle speed attribute for the 14 preceding vehicles is transmitted to the CACC controller, which then takes the average of these velocities, taking into consideration distance constraints so that the controlled vehicle does not come dangerously close to the immediate preceding vehicle or recede too far away from it. The idea behind taking the average of the velocities of the 14 preceding vehicles is to incorporate the dynamics or the behavior of the preceding vehicles ahead of time in the dynamics of the controlled vehicle. Given that ideal communication is available, taking the average of the velocities of the preceding vehicles allows the CACC controller to determine the future trajectory of the controlled vehicle, which leads to significant fuel benefits and reduction of emissions. For instance, when the lead vehicle in the platoon of vehicles stops at a traffic signal, the decrease in the velocity of this lead vehicle also causes the average of the velocities to decrease, which allows the controlled vehicle to lower its velocity and increase the relative distance between itself and the immediate preceding vehicle. This increase in relative distance gives the controlled vehicle sufficient space to maintain the same velocity instead of abruptly decelerating or accelerating in the case when communication is not available. This behavior over a complete driving cycle tends to smoothen the velocity profile of the controlled vehicle with less fluctuation in its acceleration, which has a direct correlation with fuel consumption. Hence, occurrence of this behavior over a driving cycle leads to significant fuel benefits relative to the immediate preceding vehicle's fuel consumption.

FIGURE 2 CACC Concept Sketch.

Grahic Jump LocationFIGURE 2 CACC Concept Sketch.

The experimental results were obtained using the HIL testbed. The controlled vehicle is modeled as a power-split hybrid electrical vehicle [1]. A planetary gear set connects the engine to a motor and a generator. Based on the desired speed determined by the CACC controller, the vehicle power demand is calculated and split between the engine and the battery power. The engine operating point is then calculated and sent to the powertrain research platform in real time. The results below represent the vehicle dynamics, powertrain dynamics, fuel consumption, and emissions of the controlled vehicle as well as the immediate preceding vehicle.

Figures 3 and 4 show the tracking performance of the actual engine in the powertrain research platform for the controlled and the immediate preceding vehicle, respectively. Figure 5 clearly shows that the emissions for the immediate preceding vehicle are higher for all the gases compared to the CACC controlled vehicle. Figure 6 strengthens the hypothesis that using preceding vehicles information has potential for fuel benefits by showing that there is approximately 16.9% fuel benefit for the controlled vehicle with respect to the immediate preceding vehicle in a local highway driving condition.

FIGURE 3 Vehicle and Powertrain Dynamics of the CACC Vehicle.

Grahic Jump LocationFIGURE 3 Vehicle and Powertrain Dynamics of the CACC Vehicle.

FIGURE 4 Vehicle and Powertrain Dynamics of the Immediate Preceding Vehicle.

Grahic Jump LocationFIGURE 4 Vehicle and Powertrain Dynamics of the Immediate Preceding Vehicle.

FIGURE 5 Emissions Measurements for the CACC Vehicle and the Immediate Preceding Vehicle.

Grahic Jump LocationFIGURE 5 Emissions Measurements for the CACC Vehicle and the Immediate Preceding Vehicle.

FIGURE 6 Total Fuel Consumption Comparison for the CACC vehicle and the Immediate Preceding Vehicle.

Grahic Jump LocationFIGURE 6 Total Fuel Consumption Comparison for the CACC vehicle and the Immediate Preceding Vehicle.

In the foreseeable future, connected vehicles need to operate alongside unconnected vehicles. This makes the evaluation of connected vehicles and their applications challenging. The HIL testbed can be used as a tool to evaluate the connected vehicle applications in a safe, efficient and economic fashion. The HIL testbed integrates a traffic simulation network with a powertrain research platform in real time. Any target vehicle in the traffic network can be selected so that the powertrain research platform will be operated as if it is propelling the target vehicle. The HIL testbed can also be connected to a living laboratory where actual on-road vehicles can interact with the powertrain research platform.

The authors would like to acknowledge the support by the Federal Highway Administration under grant DTFH6114H00005.

Mohd Zulkefli, M.A., J. Zheng, Z. Sun, and H. Liu. “Hybrid Powertrain Optimization with Trajectory Prediction Based on Inter-Vehicle-Communication and Vehicle-Infrastructure-Integration,” Transportation Research Part C 45 (2014): 41– 63 [CrossRef]
Hu, J., Y. Shao, Z. Sun, M. Wang, J. Bared, and P. Huang. “Integrated Optimal Eco-Driving On Rolling Terrain for Hybrid Electric Vehicle with Vehicle-Infrastructure Communication,” Transportation Research Part C 68 (2016): 228– 244. [CrossRef]
Wang, Y., andZ. Sun. “Dynamic Analysis and Multivariable Transient Control of the Power-Split Hybrid Powertrain,” IEEE/ASME Transactions on Mechatronics 20, no. 6 (Dec. 2015): 3085– 3097. [CrossRef]
Mukherjee, P. “Investigation of Cooperative Adaptive Cruise Control with Experimental Validation,” MS Thesis, University of Minnesota, 2016.
Copyright © 2016 by ASME
Topics: Vehicles , Traffic , Hardware , Roads
View article in PDF format.

References

Mohd Zulkefli, M.A., J. Zheng, Z. Sun, and H. Liu. “Hybrid Powertrain Optimization with Trajectory Prediction Based on Inter-Vehicle-Communication and Vehicle-Infrastructure-Integration,” Transportation Research Part C 45 (2014): 41– 63 [CrossRef]
Hu, J., Y. Shao, Z. Sun, M. Wang, J. Bared, and P. Huang. “Integrated Optimal Eco-Driving On Rolling Terrain for Hybrid Electric Vehicle with Vehicle-Infrastructure Communication,” Transportation Research Part C 68 (2016): 228– 244. [CrossRef]
Wang, Y., andZ. Sun. “Dynamic Analysis and Multivariable Transient Control of the Power-Split Hybrid Powertrain,” IEEE/ASME Transactions on Mechatronics 20, no. 6 (Dec. 2015): 3085– 3097. [CrossRef]
Mukherjee, P. “Investigation of Cooperative Adaptive Cruise Control with Experimental Validation,” MS Thesis, University of Minnesota, 2016.

Figures

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In