Leveraging connected vehicle platooning technology to improve the efficiency and effectiveness of train fleeting under moving blocks
Introduction
North American railroads are confronting growing demand for safe, efficient, and reliable freight and passenger transportation. Rising energy costs and competition from other transportation modes give railroads an economic incentive to optimize train operations for better fuel efficiency. Similarly, the high cost of building additional track infrastructure to increase capacity and improve reliability gives a potent inducement to railroads to enhance productivity of their existing mainlines by decreasing the headway between trains.
Facing similar efficiency and capacity demands, highway transportation has turned to intelligent transportation systems technologies. For example, adaptive cruise control (ACC) systems developed for highway vehicles have been combined with other technologies to support autonomous highway vehicle operations. Such developed systems react to surrounding terrain and traffic stream situations to control vehicle throttle and brake in a manner that aims for better fuel efficiency. More advanced versions of these systems use connected vehicle technology to form platoons of closely spaced vehicles that travel together in a coordinated manner. The reduced spacing between vehicles in a platoon decreases aerodynamic drag and consequently fuel use, at the same time increasing effective road capacity (Tsugawa et al., 2016, McAuliffe et al., 2018, Noruzoliaee et al., 2021, Sun et al., 2021).
In contrast to highways, railways have implemented driver advisory systems designed to improve fuel efficiency, by informing locomotive operators on the optimal throttle and brake commands. Systems such as LEADER from New York Air Brake have become progressively more sophisticated in their ability to support semi-autonomous adaptive train cruise control and even full autonomous (“auto pilot”) capability under test conditions in Western Australia (Vantuono, 2019). While these systems are effective in reducing fuel usage, they do not focus on minimizing train headways. These driver advisory systems also do not communicate with other trains in general. Moreover, there is no direct energy benefit from reducing train headways, because even at the minimum safe braking distance, trains are sufficiently separated that no tangible aerodynamic effect between two consecutive trains. Additionally, the ability of these systems to reduce train headway and increase line capacity is constrained by the existing wayside block signal system used to control railway traffic and maintain safe separation between trains. Advanced Positive Train Control (PTC) systems currently under development may replace this wayside system with virtual or moving blocks that allow for shorter train headways.
Conceptually, eliminating the headway inefficiency of fixed signal blocks with lengths established by the braking distance of the poorest-performing train allows trains to follow each other at minimum safe braking distances (Fig. 1). The reduced headway between trains can increase capacity on double-track mainlines, such as those found on key freight corridors with high traffic density or lines in urban areas with combinations of freight and commuter rail service (Dick, 2000, Dingler et al., 2009, Dingler et al., 2010, Dick et al., 2019, Diaz de Rivera and Dick, 2021). Reduced train headway under moving blocks can also improve capacity by minimizing the time required for meets between fleets of trains on single-track mainlines (Diaz de Rivera et al., 2020a, Diaz de Rivera et al., 2020b). Although lengthening trains (and simultaneously reducing the number of trains) may be used to improve railcar throughput on a given section of track, North American railways have pushed this strategy to its limit when implementing Precision Scheduled Railroading (Blaze, 2022). At the same time, lengthening trains has resulted in new siding extension projects to keep trains moving on the mainline, and has caused increased congestion in rail yards, where lengthening tracks is often much more difficult (Dick, 2021).
Although operating highway and rail vehicles at shorter headways can conceptually increase capacity, it is difficult to implement in practice. In actual operating settings, the lead vehicle rarely maintains a perfectly constant speed due to changes in grade, curvature, and wind speed and direction. Small fluctuations in the speed of the lead vehicle are amplified by each trailing vehicle as they react and attempt to maintain a minimum required headway. Tests of highway vehicles with ACC have shown that platoons of vehicles exhibit a “harmonic” effect where the headway between subsequent vehicles rapidly expands and shrinks as following vehicles attempt to match the speed of the lead vehicle. Connecting vehicles to share throttle and brake commands can allow for new control algorithms that improve fuel efficiency and the ability of the vehicles to maintain a constant minimum headway. This research aims to learn more about these vehicle control algorithms and how they can be used to solve similar problems when fleets of closely following trains are operated under advanced PTC systems.
The objective of this research is to leverage existing and developing connected highway vehicle platooning technology to improve the efficiency and effectiveness of fleeting trains at minimum headways under moving blocks. The research aims to 1) better understand how closely following trains respond to different throttle and brake control algorithms and 2) using insights gained from connected automobile and truck platooning technology, develop improved train control algorithms that allow railway operators to balance fuel efficiency and train throughput. It is worth noting that interfacing with or integrating a train fuel consumption minimizer is not a part of this research. In order to do so, it will be important to first develop an effective train headway minimization controller for moving blocks. Additionally, most North American rail routes for heavy haul freight where moving blocks make sense are often over-capacitated, suggesting a strongly priority placed on headway minimization, which can help reduce the need for new track infrastructure, over fuel use reduction.
To achieve the objectives of the paper, the research uses simulations to better understand the baseline headway and fuel efficiency implications of this type of operation. We develop and simulate improved “train following” control algorithms to better manage train fleeting operations, adapting existing approaches to highway vehicle platooning where applicable. The results allow industry practitioners to develop improved locomotive driver advisory and semi-autonomous adaptive train cruise control systems for train fleeting operations under moving blocks, and allow railroad operators to make more informed decisions based on the potential fuel efficiency and capacity benefits of these systems.
The research scope is limited to two different communication topology scenarios: a basic scenario where the following trains in a fleet only receive information about the position of the train ahead at discrete intervals, and a scenario with train-to-train communications of throttle and brake actions to the following trains. The latter scenario may allow the following trains to anticipate changes in the speed of the lead train and make more efficient and timely control decisions that preserve both headway and fuel efficiency. While such train-to-train communications in the context of North American freight rail operations remain difficult today, the communication network provided by future advanced PTC systems with moving blocks are expected to facilitate this type of coordination in the future.
The remainder of the paper is organized as follows. Section 2 reviews the existing literature on railway train control algorithms, car-following models with and without the connected vehicle technology, and truck platooning control. Section 3 describes train movement dynamics and the simulation model developed to evaluate the train following control algorithms. Section 4 details the development and formulation of five control algorithms, including three that are adapted from the highway literature and two proposed by ourselves. Section 5 outlines the train following simulation experiments, and presents and compares the performance of the five control algorithms. Section 6 summarizes the conclusions of this research.
Section snippets
Literature review
In pursuit of the paper’s research objective, we conducted an extensive literature review to determine the current state of knowledge regarding control algorithms for closely following trains and highway vehicles. The following subsections summarize the key findings of the review, based on which we identify the literature gap and present the contributions of our present work.
Train movement dynamics and motivation for considering control algorithms
The dynamics of train in a fleet of n trains can be expressed as follows:where , , and are the position, speed, and mass of train at time , respectively. is the enforced traction or braking force, represents the aerodynamic and basic resistances, and denotes net of other external resistance forces due to gradient of track and curve alignment (Gao et al., 2016). From a technical perspective, the objective of designing a
Following train control algorithm development
This section describes the control algorithms to investigate in the context of train fleeting, by leveraging the literature on highway vehicle following and train following. While most of the controllers considered are linear, an alternative approach could be using non-linear controllers as for highway vehicles (Coppola et al., 2022, Zuo et al., 2022). However, it is expected that controllers that explicitly consider non-linear train dynamics are not necessary for train control applications.
Simulated performance of following train control algorithms
To evaluate each of the proposed control algorithms, a battery of tests was developed based on a factorial design. This was done by testing values for scenario parameters which seemed most likely to produce undesirable results from the control algorithms, and then iteratively determining which of these values consistently resulted in undesirable behavior. By focusing on these values, the number of simulations needed to evaluate control algorithm behavior was greatly reduced, especially because
Conclusion
Directly adapting highway vehicle platooning controllers to the heavy-haul freight and passenger railway domain is difficult due to the orders of magnitude difference in highway and rail vehicle performance. However, highway controllers do suggest families of control laws to adapt to the train following problem. The simulation results suggest that certain families of control laws are better than others at managing train separation and fuel consumption within train fleets. Certain controllers
CRediT authorship contribution statement
Pooria Choobchian: Methodology, Investigation, Writing – original draft, Writing – review & editing. Geordie Roscoe: Methodology, Investigation, Writing – original draft, Writing – review & editing, Visualization. Tyler Dick: Conceptualization, Methodology, Investigation, Writing – original draft, Writing – review & editing, Supervision. Bo Zou: Conceptualization, Methodology, Investigation, Writing – original draft, Writing – review & editing, Supervision. Daniel Work: Conceptualization,
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgement
This study was supported by the Federal Railroad Administration of the U.S. Department of Transportation (USDOT). The opinions expressed are solely those of the authors, and do not necessarily represent those of USDOT.
References (57)
- et al.
The key principles of optimal train control—Part 1: Formulation of the model, strategies of optimal type, evolutionary lines, location of optimal switching points
Transportation Research Part B: Methodological.
(2016) - et al.
Stability Analysis of Train Following Model with Multiple Communication Delays
IFAC Proceedings Volumes.
(2013) - et al.
Illustrating the implications of moving blocks on railway traffic flow behavior with fundamental diagrams
Transportation Research Part C: Emerging Technologies.
(2021) - et al.
Train Speed Control in Moving-Block Railway Systems: An Online Adaptive PD Controller Design
IFAC Proceedings Volumes.
(2013) - et al.
String stability for vehicular platoon control: Definitions and analysis methods
Annual Reviews in Control.
(2019) - et al.
Comparative analysis of algorithms and models for train running simulation
Journal of Rail Transport Planning & Management.
(2014) - et al.
Enhanced cooperative car-following traffic model with the combination of V2V and V2I communication
Transportation Research Part B: Methodological.
(2016) - et al.
Platoon based cooperative driving model with consideration of realistic inter-vehicle communication
Transportation Research Part C: Emerging Technologies.
(2016) - et al.
Coordinated cruise control for high-speed train movements based on a multi-agent model
Transportation Research Part C: Emerging Technologies.
(2015) - et al.
Modeling cooperative and autonomous adaptive cruise control dynamic responses using experimental data
Transportation Research Part C: Emerging Technologies.
(2014)
Truck platooning in the US national road network: A system-level modeling approach
Transportation Research Part E: Logistics and Transportation Review.
Dissipation of stop-and-go waves via control of autonomous vehicles: Field experiments
Transportation Research Part C: Emerging Technologies.
Influence of connected and autonomous vehicles on traffic flow stability and throughput
Transportation Research Part C: Emerging Technologies.
Cooperative Control for Train Headway Adjustment in Railway Traffic
IFAC Proceedings Volumes.
A distributionally robust stochastic optimization-based Model Predictive Control with distributionally robust chance constraints for Cooperative Adaptive Cruise Control under uncertain traffic conditions
Transportation Research Part B: Methodological
Rolling horizon stochastic optimal control strategy for ACC and CACC under uncertainty
Transportation Research Part C: Emerging Technologies.
Distributed model predictive control approach for cooperative car-following with guaranteed local and string stability
Transportation Research Part B: Methodological.
Eco-driving control architecture for platoons of uncertain heterogeneous nonlinear connected autonomous electric vehicles
IEEE Transactions on Intelligent Transportation Systems.
Improving railway operational efficiency with moving blocks, train fleeting and alternative single-track configurations
Transportation Research Record: Journal of the Transportation Research Board.
Potential for moving blocks and train fleets to enable faster train meets on single-track rail corridors
Journal of Transportation Engineering, Part A: Systems.
Precision Scheduled Railroading and the Need for Improved Estimates of Yard Capacity and Performance Considering Traffic Complexity
Transportation Research Record: Journal of the Transportation Research Board.
Relative capacity and performance of fixed and moving block control systems on North American freight railway lines and shared passenger corridors
Transportation Research Record: Journal of the Transportation Research Board.
Effects of communications-based train control and electronically controlled pneumatic brakes on railroad capacity
Transportation Research Record: Journal of the Transportation Research Board.
Method for validating the train motion equations used for passenger rail vehicle simulation
Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit.
Cited by (2)
Headway compression oriented trajectory optimization for virtual coupling of heavy-haul trains
2024, Control Engineering Practice
- 1
Pooria Choobchian and Geordie Roscoe are co-first authors.