Kuilin Zhang | Michigan Technological University (original) (raw)

Papers by Kuilin Zhang

Transportation Research Record, Nov 1, 2022

As freight rail transportation in the U.S.A. increasingly concentrates on high-volume and longer ... more As freight rail transportation in the U.S.A. increasingly concentrates on high-volume and longer distance train movements, shippers in rural areas with fragmented freight volumes, lighter line densities, and shorter distances have struggled to maintain the rail share of their shipments. The forest products industry in the mostly rural Lake Superior region is no exception and has seen a transition toward truck transportation for its main raw material, logs. The forest industry has attributed the phenomenon to reduced rail access and increased rail rates, and asked us to use their log shipment and rail operations network data to explore if (1) improved access through shared sidings, re-opened sidings, or both, and (2) rail rate reductions would encourage a modal shift from trucks back to rail. We developed four integer/mixed-integer programming models and several scenarios to research the problem. We concentrated on (1) allowing log consolidation from multiple companies at rail sidings and (2) rail rate reductions with and without volume thresholds. Our results showed that additional flexibility through shared rail sidings, re-opening of unused/closed sidings, or both, has minimal impact on increasing the rail modal share. Rail rate discounts with or without minimum volume thresholds were able to encourage a modal shift, but growing volume through rate reductions may not be attractive from the rail service provider perspective. While the results were not encouraging, our models offer the industry a data-based approach for further investigations, such as the impacts of potential shifts to a shortline cost model, or targeted analysis of specific mills that showed greater promise for a modal shift.

Transportation Research Record, May 18, 2022

Real-time control of a fleet of Connected and Automated Vehicles (CAV) for Cooperative Adaptive C... more Real-time control of a fleet of Connected and Automated Vehicles (CAV) for Cooperative Adaptive Cruise Control (CACC) is a challenging problem concerning time delays (from sensing, communication, and computation) and actuator lag. This paper proposes a real-time predictive distributed CACC control framework that addresses time delays and actuator lag issues in the real-time networked control systems. We first formulate a Kalman Filter-based real-time current driving state prediction model to provide more accurate initial conditions for the distributed CACC controller by compensating time delays using sensing data from multi-rate onboard sensors (e.g., Radar, GPS, wheel speed, and accelerometer), and status-sharing and intent-sharing data in BSM via V2V communication. We solve the prediction model using a sequential Kalman Filter update process for multi-rate sensing data to improve computational efficiency. We propose a real-time distributed MPC-based CACC controller with actuator lag and intent-sharing information for each CAV with the delay-compensated predicted current driving states as initial conditions. We implement the real-time predictive distributed CACC control algorithms and conduct numerical analyses to demonstrate the benefits of intent-sharing-based distributed computing, delay compensation, and actuator lag consideration on string stability under various traffic dynamics.

Transportation Research Record, Jun 8, 2022

Cooperative adaptive cruise control (CACC) is one of the popular connected and automated vehicle ... more Cooperative adaptive cruise control (CACC) is one of the popular connected and automated vehicle (CAV) applications for cooperative driving automation with combined connectivity and automation technologies to improve string stability. This study aimed to derive the string stability conditions of a CACC controller and analyze the impacts of CACC on string stability for both a fleet of homogeneous CAVs and for heterogeneous traffic with human-driven vehicles (HDVs), connected vehicles (CVs) with connectivity technologies only, and autonomous vehicles (AVs) with automation technologies only. We mathematically analyzed the impact of CACC on string stability for both homogeneous and heterogeneous traffic flow. We adopted parameters from literature for HDVs, CVs, and AVs for the heterogeneous traffic case. We found there was a minimum constant time headway required for each parameter design to ensure stability in homogeneous CACC traffic. In addition, the constant time headway and the length of control time interval had positive correlation with stability, but the control parameter had a negative correlation with stability. The numerical analysis also showed that CACC vehicles could maintain string stability better than CVs and AVs under low HDV market penetration rates for the mixed traffic case.

Transportation Research Part C: Emerging Technologies

Transportation Research Board 98th Annual MeetingTransportation Research Board, 2018

Motivated by Connected and Automated Vehicle (CAV) technologies, this paper proposes a data-drive... more Motivated by Connected and Automated Vehicle (CAV) technologies, this paper proposes a data-driven optimization based Model Predictive Control (MPC) modeling framework for real-time automatically controlling ECO Approach and Departure (ECO-AND) under uncertain traffic conditions. The proposed data-driven optimization based MPC modeling framework aims to improve the safety, energy efficiency, driving comfort, and robustness of the ECO-AND longitudinal automated driving under uncertain traffic conditions by using Vehicle-to-Infrastructure (V2I) and Vehicle-to-Vehicle (V2V) data. Through an online learning-based driving dynamics prediction model, the authors predict a set of uncertain driving states of the preceding vehicles in front of the controlled CAV. With the predicted driving states of the preceding vehicles, the authors solve a constrained Finite-Horizon Optimal Control problem considering the trade-off between energy consumption and driving efficiency to predict a set of uncertain driving states of the controlled CAV as ECO-driving references. To obtain the optimal acceleration or deceleration commands for the CAV with the set of ECO-driving references, the authors formulate a Distributionally Robust Stochastic Optimization (DRSO) model (i.e. a special case of data-driven optimization models under moment bounds) with Distributionally Robust Chance Constraints (DRCC) that explicitly counts for location-based intersection traffic signal control constraints as well as safe driving constraints. To solve the minimax program of the DRSO-DRCC model, the authors reformulate a relaxed dual problem as a Semidefinite Program (SDP) based on the strong duality theory and the Semidefinite Relaxation technique. In addition, the authors propose a solution algorithm to solve the relaxed SDP problem. The authors design experiments to demonstrate that the proposed model and conduct computational analyses to validate the efficiency of the proposed algorithm for solving the DRSO-DRCC model for real-time automated Eco-driving applications

Traffic And Transportation Studies (2002), 2002

This study aims at illustrating the relationship between traffic parameters and energy/fuel consu... more This study aims at illustrating the relationship between traffic parameters and energy/fuel consumption., using real-world data. The dataset is a 500 m stretch of a Californian freeway during rush hours. Each vehicle captured by cameras on that stretch is characterized by a subsecond position and speed. This speed vs. time trace is then used in Autonomie, a powertrain simulation tool, to compute energy consumption. The analysis shows a positive correlation between traffic density and energy consumption.

This paper proposes a data-driven dynamic route choice model to understand traveler’s routing beh... more This paper proposes a data-driven dynamic route choice model to understand traveler’s routing behavior in a time-dependent network under uncertainty using connected vehicle trajectory data over many days. Different from existing efforts on stochastic route choice models using a random term with a given distribution, this paper directly uses connected vehicle trajectory data over many days without knowing the underlying distribution in a data-driven stochastic optimization model. Specifically, the authors apply a Bayesian risk formulation for parametric underlying distributions that optimizes a risk measure taken with respect to the posterior distribution estimated from the connected vehicle trajectory data. Two risk measures (i.e. Value-at-Risk and Conditional Value-at-Risk) of the travel time uncertainty are considered in the proposed data-driven dynamic route choice model. Based on the risk measures, the proposed model allows a flexible choice on the risk preferences of individual users (i.e. from risk-neutral to risk-averse). To test the data-driven dynamic route choice model in a large network, the authors implement the model in Southeast Michigan using a high-resolution (i.e. 0.1 seconds) trajectory dataset of connected vehicles from the Safety Pilot Model Deployment (SPMD) project over many days

2017 IEEE Power & Energy Society Innovative Smart Grid Technologies Conference (ISGT), 2017

Real-time monitoring has become a critical part of distribution network operation that enhances t... more Real-time monitoring has become a critical part of distribution network operation that enhances the control and automation capabilities as metering technologies evolve. The metering infrastructure has further extended from feeder head of a substation throughout the entire feeder loads. Despite installations of “smart” meters and related recording devices are increasing rapidly, the measurable area does not reach the ideal status that each load should be installed with a “smart” meter to constantly observe the electric information. This paper proposes a statistical approach to correlate estimated occupancy datasets of buildings with smart meter building associated with a partially observable distribution feeder. This study includes a sensitivity analysis of occupancy how it can affect load consumptions with or without the temperature load. This also creates a general load profile for other unmetered buildings within the distribution feeder where the feeder head is assumed to be obser...

Sustainability, 2020

Rail car availability and the challenges associated with the seasonal dynamics of log movements h... more Rail car availability and the challenges associated with the seasonal dynamics of log movements have received growing attentions in the Lake Superior region of the US, as a portion of rail car fleet is close to reaching the end of its service life. This paper proposes a data-driven study on the rail car peaking issue to explore the fleet of rail cars dedicated to being used for log movements in the region, and to evaluate how the number of cars affects both the storage need at the sidings and the time the cars are idled. This study is based on the actual log scale data collected from a group of forest companies in cooperation with the Lake State Shippers Association (LSSA). The results of our analysis revealed that moving the current log volumes in the region would require approximately 400–600 dedicated and shared log cars in ideal conditions, depending on the specific month. While the higher fleet size could move the logs as they arrive to the siding, the lower end would nearly el...

2015 IEEE Transportation Electrification Conference and Expo (ITEC), 2015

Electric vehicles are increasingly being adopted due to environmental awareness and competitive t... more Electric vehicles are increasingly being adopted due to environmental awareness and competitive technical performance and reducing prices. Their research and development has sometimes relied on the use of standard driving cycles. However, these cycles cannot reproduce the variations of traffic flow in real world. That is why in this paper we develop a software tool able to analyze real-life driving cycles for electric vehicles. To do so, a driving trajectory process tool is used to obtain large data for vehicles driving in the same stretch of highway. To show the performance of the developed tool, sample cycles are analyzed and simulated for electric vehicles automatically in the REV-Cycle (Real Electric Vehicle Cycle analyzer) software presented.

This paper presents a theoretically sound time-dependent stochastic user equilibrium (TDSUE) traf... more This paper presents a theoretically sound time-dependent stochastic user equilibrium (TDSUE) traffic assignment model and its simulation-based solution algorithm within a probit-based path choice decision framework. The TDSUE problem, which aims to find time-dependent SUE path flows, is typically considered as a fixed point problem; it is reformulated as an equivalent gap function-based nonlinear optimization problem, and then solved by a column generation-based solution framework which embeds (i) a simulation-based dynamic network loading model to capture traffic dynamics and determine experienced time-dependent path travel disutilities and temporal and spatial path correlations for a given path flow pattern; (ii) a projected gradient-based descent direction method to solve the restricted SUE problem defined by a subset of feasible paths; and (iii) a time-dependent least-cost path algorithm to generate paths to augment the feasible path set. This nonlinear optimization reformulatio...

Transportation Research Part B-methodological, 2021

For energy-efficient Connected and Automated Vehicle (CAV) Eco-driving control on signalized arte... more For energy-efficient Connected and Automated Vehicle (CAV) Eco-driving control on signalized arterials under uncertain traffic conditions, this paper explicitly considers traffic control devices (e.g., road markings, traffic signs, and traffic signals) and road geometry (e.g., road shapes, road boundaries, and road grades) constraints in a data-driven optimization-based Model Predictive Control (MPC) modeling framework. This modeling framework uses real-time vehicle driving and traffic signal data via Vehicle-to-Infrastructure (V2I) and Vehicle-to-Vehicle (V2V) communications. In the MPC-based control model, this paper mathematically formulates location-based traffic control devices and road geometry constraints using the geographic information from High-Definition (HD) maps. The location-based traffic control devices and road geometry constraints have the potential to improve the safety, energy, efficiency, driving comfort, and robustness of connected and automated driving on real ...

Technological advances in unmanned aerial vehicle (UAV) technologies continue to enable these too... more Technological advances in unmanned aerial vehicle (UAV) technologies continue to enable these tools to become easier to use, more economical, and applicable for transportation-related operations, maintenance, and asset management while also increasing safety and decreasing cost. This Phase 2 project continued to test and evaluate five main UAV platforms with a combination of optical, thermal, and lidar sensors to determine how to implement them into MDOT workflows. Field demonstrations were completed at bridges, a construction site, road corridors, and along highways with data being processed and analyzed using customized algorithms and tools. Additionally, a cost-benefit analysis was conducted, comparing manual and UAV-based inspection methods. The project team also gave a series of technical demonstrations and conference presentations, enabling outreach to interested audiences who gained understanding of the potential implementation of this technology and the advanced research that MDOT is moving to implementation. The outreach efforts and research activities performed under this contract demonstrated how implementing UAV technologies into MDOT workflows can provide many benefits to MDOT and the motoring public; such as advantages in improved cost-effectiveness, operational management, and timely maintenance of Michigan's transportation infrastructure 17.

ICC 2020 - 2020 IEEE International Conference on Communications (ICC), 2020

Effective clustering is vital to mitigate routing scalability and reliability issues in heterogen... more Effective clustering is vital to mitigate routing scalability and reliability issues in heterogeneous vehicular networks. In this paper, we propose an adaptive clustering scheme to maximize the cluster stability in vehicular networks. The scheme uses the predicted driving behavior of vehicles over a time horizon to maximize the clusters’ lifetime. To this end, we first define the stability degree of vehicles by exploiting the unique aspects of vehicular environments. We then formulate the clustering problem as an optimization problem, which is used within a rolling horizon framework in the cluster formation process. Our scheme is based on a heterogeneous vehicular network architecture, which allows the coexistence of dedicated short-range communication and cellular network for vehicular communications. The simulation results demonstrate that our scheme significantly outperforms alternative clustering algorithms in terms of the overall clusters’ lifetime under different traffic condi...

This dissertation aims at developing models and algorithms for the dynamic stochastic user equili... more This dissertation aims at developing models and algorithms for the dynamic stochastic user equilibrium (DSUE) problem in multidimensional transportation networks. These models form the basis of decision-support tools for the planning and design of advanced operational strategies to reduce traffic congestion and enhance system sustainability. The central contribution is a behaviorally flexible time-dependent stochastic user equilibrium (TDSUE) model, which integrates probit-based discrete choice models of user path decisions in a simulation-based dynamic assignment framework. This fixed-point problem is reformulated as an equivalent gap function-based nonlinear optimization problem, and solved by a column-generation framework with a vehicle-based implementation. The approach incorporates newly developed methods to specify and compute the variance-covariance matrix for stochastic dynamic path choice models, thereby addressing a fundamental challenge in capturing spatial and temporal c...