ARC Researchers at 2018 ASME DSCC
Dynamic Systems and Control Conference
(September 30 - October 3 at Atlanta, Georgia)

ARC researchers (principal investigators in bold) will be presenting their latest research developments. Below are their papers which may include non-ARC funded research (paper titles in bold are ARC funded).

Monday, October 1, 2018
Time, Room Session, Paper Title, Author, Abstract
10:00 AM - 12:00 PM
1-3 Advances in Control Design Methods (Contributed Session)
Stochastic Policies for Online Computation Triggering in Powertrain Control
  Mr. Kuan Liu, Graduate Student Research Assistant, University of Michigan -- Ann Arbor
Dr. Yue Yun Wang, Research Engineer, General Motors R&D Ctr
Dr. Ibrahim Haskara, Staff Research Engineer, General Motors R&D Ctr
Dr. Chenfang Chang, Staff Research Engineer, General Motors R&D Ctr
Prof. Anouck Girard, Assistant Professor, University of Michigan
Prof. Ilya Kolmanovsky, Professor, The University of Michigan, Ann Arbor
  With the rapid growth in the amount of computations that need to be performed by modern electronic control units and in the complexity of the algorithms, there is a pressing need to develop approaches to reduce chronometric loading in automotive vehicles. This paper presents a cyberphysical systems framework for the development of stochastic optimal decision policies that trigger the computations online. For a specific case study, we consider online triggering of the computations involved in obtaining a linearized model in a setting when such a linearized model is required by control or estimation algorithms. The objective is to define a policy for triggering the linearization that balances the average model accuracy (or expected closed loop performance) with the average computational cost. The problem is formulated as a stochastic optimal control problem and solved using stochastic dynamic programming (SDP). The approach is described, then illustrated with three examples, a pendulum, a turbocharged diesel engine, and a turbocharged spark ignition engine, that illustrate the trade-off between the computational cost and expected linearized model accuracy.
10:00 AM - 12:00 PM Kennesaw 1-6 Path Planning and Motion Control (Contributed Session)
Trust-Based Run-time Verification for Multi-Quadrotor Motion Planning with A Human-in-the-Loop
  Mr. Maziar Fooladi Mahani, Student, Clemson University
Prof. Yue Wang, Professor, Clemson University
  In this paper, we propose a trust-based run-time verification (RV) framework for deploying multiple quadrotors with a human-in-the-loop (HIL). By bringing together approaches from run-time verification, trust-based decision-making, human-robot interaction (HRI), and hybrid systems, we develop a unified framework that is capable of integrating human cognitive skills with autonomous capabilities of multi-robot systems to improve system performance and maximize the intuitiveness of HRI. We also use the automata theoretic approaches to generate plans for a quadrotor working in a partially-known environment by automatic synthesis of controllers enforcing specifications given in temporal languages. On top of the RV framework, we utilize a trust-based decision module as the key component in forming the HRI, designed to maintain the system performance. To demonstrate the feasibility of the framework, it is implemented using a bisimilar control design together with simulated case studies.
10:00 AM - 12:00 PM Techwood 1-10 Multi-Agent and Networked Systems (Contributed Session)
Human-Robot Trust Integrated Task Allocation and Symbolic Motion Planning for Heterogeneous Multi-robot Systems
  Mr. Huanfei Zheng, Ph.D. student, Clemson University
Mr. Zhanrui Liao, Ph.D. student, Clemson University,
Prof. Yue Wang, Professor, Clemson University
  This paper presents a human-robot trust integrated task allocation and motion planning framework for multi-robot systems (MRS) in performing a set of parallel subtasks. Parallel subtask specifications are conjuncted with MRS to synthesize a task allocation automaton. Each transition of the task allocation automaton is associated with the total trust value of human in corresponding robots. A dynamic Bayesian network (DBN) based human-robot trust model is constructed considering individual robot performance, safety coefficient, human cognitive workload and overall evaluation of task allocation. Hence, a task allocation path with maximum encoded human-robot trust can be searched based on the current trust value of each robot in the task allocation automaton. Symbolic motion planning (SMP) is implemented for each robot after they obtain the sequence of actions. The task allocation path can be intermittently updated with this DBN based trust model. The overall strategy is demonstrated by a simulation with 5 robots and 3 parallel subtask automata.
01:30 PM - 03:30 PM Inman 2-2 Automotive Dynamics and Emerging Powertrain Technologies (Invited Session)
Predictively Coordinated Vehicle Acceleration and Lane Selection Using Mixed Integer Programming
  Mr. Robert Dollar, Research Assistant, Clemson University
Prof. Ardalan Vahidi, Professor, Clemson University
  Autonomous vehicle technology provides the means to optimize motion planning beyond human capacity. In particular, the problem of navigating multi-lane traffic optimally for trip time, energy efficiency, and collision avoidance presents challenges beyond those of single-lane roadways. For example, the host vehicle must simultaneously track multiple obstacles, the drivable region is non-convex, and automated vehicles must obey social expectations. Furthermore, reactive decision-making may result in becoming stuck in an undesirable traffic position. This paper presents a fundamental approach to these problems using model predictive control with a mixed integer quadratic program at its core. Lateral and longitudinal movements are coordinated to avoid collisions, track a velocity and lane, and minimize acceleration. Vehicle-to-vehicle connectivity provides a preview of surrounding vehicles’ motion. Simulation results show a 79% reduction in congestion-induced travel time and an 80% decrease in congestion-induced fuel consumption compared to a rule-based approach.
04:00 PM - 06:00 PM Techwood 1-39 Advances in Robotics II (Contributed Session)
Trust-Based Impedance Control Strategy for Human-Robot Cooperative Manipulation
  Mr. Behzad Sadrfaridpour, Ph.D. student, Clemson University
Mr. Zhanrui Liao, Ph.D. student, Clemson University
Mr. Maziar Fooladi Mahani, Student, Clemson University
Prof. Yue Wang, Professor, Clemson University
  A trust-based switching impedance control strategy for human-robot cooperative manipulation is proposed. The robot behavior switches between the proactive and reactive modes based on the estimated trust of human in the human-robot cooperative system. The robot estimates the human desired trajectory A history-based probabilistic trust model for human-robot cooperative manipulation tasks is proposed. Trust is modeled as a dynamic Bayesian network. In the proactive mode, the robot estimates the human desired trajectory and plans accordingly while in the reactive mode, robot only reacts to the human input. A simulation of the trust-based switching impedance control strategy is presented.
04:00 PM - 06:00 PM Inman 2-10 Control and Optimization of Connected and Automated Ground Vehicles (Invited Session)
Optimization of Energy-Efficient Speed Profile for Electrified Vehicles
  Mr. Hadi Abbas, Graduate Student Research Assistant, University of Michigan
Prof. Youngki Kim, Assistant Professor, University of Michigan
Dr. Jason Siegel, Assistant Research Scientist, Univ Of Michigan
Dr. Denise Rizzo, Senior Research Mechanical Engineer, U.S. Army TARDEC
  This paper presents a study of energy-efficient operation of vehicles with electrified powertrains leveraging route information, such as road grades, to adjust the speed trajectory. First, Pontryagin's Maximum Principle (PMP) is applied to derive necessary conditions and to determine the possible operating modes. The analysis shows that only 5 modes are required to achieve minimum energy consumption; full propulsion, cruising, coasting, full regeneration, and full regeneration with conventional braking. The minimum energy consumption problem is reformulated and solved in the distance domain using Dynamic Programming to optimize speed profiles. A case study is shown for a light weight military robot including road grades. For this system, a tradeoff between energy consumption and trip time was found. The optimal cycle uses 20% less energy for the same trip duration, or could reduce the travel time by 14% with the same energy consumption compared to the baseline operation.
Tuesday, October 2, 2018
Time, Room Session, Paper Title, Author, Abstract
01:30 PM - 03:30 PM Spring 1-29 Energy Systems (Contributed Session)
Comparison Of Individual-Electrode State Of Health Estimation Methods For Lithium Ion Battery
  Mr. Suhak Lee, Graduate Student Research Assistant, University of Michigan
Dr. Jason Siegel, Assistant Research Scientist, Univ Of Michigan
Prof. Anna Stefanopoulou, Professor, Univ Of Michigan
Mr. Jang-Woo Lee, Senior Engineer, Samsung SDI Co., Ltd.
Dr. Tae-Kyung Lee, Vice President, Samsung SDI Co., Ltd.
  It is essential to understand the state-of-health (SOH) of the individual electrodes to avoid accelerating degradation of Li-ion battery. Electrode SOH can be quantified based on estimating the capacity and the utilization range of each electrode. Here, we introduce two methods: i) voltage fitting (VF) and ii) peak alignment (PA), and compare their ability to estimate the electrode SOH parameters. Both methods assume the half-cell open-circuit potentials (OCPs) are invariant functions of the stoichiometric states as the cell ages, which can make the accuracy of the electrode parameter estimation vulnerable to degradation mechanisms that would cause changes in the half-cell OCP curves. This hypothesis is verified experimentally by applying the two methods to aged cells cycled at high temperature. A discernible misalignment of the peaks is observed in the differential voltage curve from the VF method indicating the electrode SOH parameter estimates are incorrect. Therefore, lower voltage error of the cell and accurate cell capacity estimate do not necessarily yield better estimation accuracy for the electrode SOH parameters.
01:30 PM - 03:30 PM Inman 1-30 Automotive Systems (Contributed Session)
Energy-efficient Control Approach for Automated HEV and BEV with Short-horizon Preview Information
  Dr. Jinwoo Seok, Postdoctoral research fellow, University of Michigan
Dr. Yan Wang, Technical expert, Ford Motor Company
Dr. Dimitar Filev, Executive Technical Leader, Ford Motor Company
Prof. Ilya Kolmanovsky, Professor, The University of Michigan, Ann Arbor
Prof. Anouck Girard, Associate Professor, University of Michigan
  The interest in autonomous electrified vehicles has rapidly increased in recent years. In this paper, a speed trajectory generation method based on speed and road grade preview over a short horizon is presented, to improve energy efficiency of the autonomous power split HEV and BEV. The speed trajectory is generated by minimizing the required cumulative over the the horizon wheel power, including regenerative power. We confirm by simulations that the generated speed profiles improve MPG (or MPGe) even though only a short horizon preview information is used with small deviation permitted from the nominal speed profile.
01:30 PM - 03:30 PM Inman 1-30 Automotive Systems (Contributed Session)
A Comprehensive Framework for Simulating Dynamics of an Off-road Vehicle in Uconstructed Environments
  Mr. Shahab Karimi, Graduate research/teaching assistant, Clemson University
Prof. Ardalan Vahidi, Professor, Clemson University
Dr. Paramsothy Jayakumar, U.S. Army Tardec
  Vehicle dynamics analysis becomes more demanding for off-road vehicles' mobility in unconstructed environments. Significant vehicle orientation changes, extreme changes in ground elevation, and uneven ground profiles at tire-road contact regions, etc. must be taken into account. In addition, the simulation computations should strike a balance between the speed and the accuracy of the results. In this paper, a model with fourteen degrees of freedom is used for vehicle dynamics analysis. Integrated within the model, a comprehensive tire model and a system of instantaneous rotation matrices are programmed to address the effect of more extreme ground profiles on the vehicle dynamics. Additionally, an iterative algorithm is developed to explore and determine the tire-road contact point. The results of simulation for two random scenarios are validated versus a commercial vehicle dynamics software showing consistency of results.
10:00 AM - 12:00 PM Piedmont 2-4 Modeling and Management of Power Systems (Invited Session)
Modeling Li-ion Battery Thermal Runaway Using A Three Section Thermal Model
  Mr. Ting Cai, PhD Candidate, University of Michigan,
Prof. Anna Stefanopoulou, Professor, Univ Of Michigan
Dr. Jason Siegel, Assistant Research Scientist, Univ Of Michigan
  This paper presents a model describing lithium-ion battery thermal runaway triggered by an internal short. The model predicts temperature and heat generation from the internal short circuit and side reactions using a three-section model. The three sections correspond to the core, middle, and surface layers. At each layer, the temperature-dependent heat release and progression of the three major side reactions are modeled. A thermal runaway test was conducted on a 4.5 Ah nickel manganese cobalt oxide pouch cell, and the temperature measurements are used for model validation. The proposed reduced order model based on three sections can balance the computational speed with the model complexity required to predict the fast core temperature evolution and slower surface temperature growth. The model shows good agreement with the experimental data, and it will be further improved with formal tuning in a follow-up effort to enable early detection of thermal runway induced by internal short.
10:00 AM - 12:00 PM Inman 1-23 Modeling and Control of IC Engines and Aftertreatment Systems (Contributed Session)
Optimal Exhaust Valve Opening Control For Fast Aftertreatmnet Warm Up in Diesel Engines
  Dr. Rasoul Salehi, Postdoc, University of Michigan
Prof. Anna Stefanopoulou, Professor, Univ Of Michigan
  This paper proposes to optimally adjust the exhaust valve opening (EVO) timing for faster selective catalytic reduction (SCR) aftertreatment system warm-up during the cold start phase of the federal test procedure (FTP). Early termination of the power stroke by EVO timing advance increases the engine exhaust gas temperature. It, on the other hand, causes exhaust flow rate reduction that decreases the coefficient of the heat transfer from the exhaust gas to the catalyst. The competing effects along with the fuel consumption increase associated with early EVO need careful consideration and the optimal EVO timing is a load-dependent balance of all these effects. This careful balance is achieved in this paper by dynamic programing (DP). Specifically, the minimum time to light-off (TTL) is formulated and applied to the cold phase of the FTP. A high fidelity detailed and verified engine and aftertreatment model is effectively simplified to enable utilizing computationally expensive DP optimization algorithm. Optimization results indicate that advancing the EVO reduces the TTL for the SCR catalyst from 659 s to 500 s, a 24% reduction. This fastest possible increase in the SCR temperature is shown to be with an expense of 4.1% increase in the fuel consumption. The results are dependent to the target light-off temperature and the load profile. Assuming a specific light-off temperature and the FTP, possible rule-based scenarios for on-line optimization are discussed.
Wednesday, October 3, 2018
Time, Room Session, Paper Title, Author, Abstract
10:00 AM - 12:00 PM Kennesaw 1-31 Mechatronics II (Contributed Session)
An Atmospheric Energy Harvester System - Linear Model and Test
  Ms. Sneha Ganesh, Graduate Student, Clemson University
Dr. Todd Schweisinger, Senior Lecturer, Clemson University
Dr. John Wagner, Professor, Clemson Univ
  Energy harvesters are steadily gaining popularity as a power source for microelectronic circuits, particularly in wireless sensor nodes and autonomous devices. Energy harvesting from small temperature and/or pressure variations, coupled with an appropriate energy storage unit, can generate sufficient electric power to operate low power electronics. Ongoing research in this area seeks to improve the power capacity and conversion efficiencies of such systems. In this project, a phase change vapor based atmospheric energy harvester with an electromechanical power transformer has been developed. An ethyl chloride fluid system converts the pressure generated, in response to nominal environmental changes, into usable electric power through a mechanical driveline-spring unit and attached DC generator. Published numerical results have indicated 9.6 mW power generation capacity over a 24 hour period for a low frequency sinusoidal temperature input with ±1°C variation at standard pressure. A prototype electromechanical unit was fabricated and experimentally tested; 30 mW electric power for a resistance load was recorded using an emulated input corresponding to 50 bidirectional cyclic atmospheric variations (~175 hour period). Linearized models were derived to help evaluate the system’s transient characteristics and these mathematical results agreed favorably with the experimental behavior.
10:00 AM - 12:00 PM Lenox 1-4 Estimation and Identification (Contributed Session)
Behavior Inference from Bio-logging Sensors: A Systematic Approach for Feature Generation, Selection and State Classification
  Mr. Ding Zhang, Student, University of Michigan
Prof. K. Alex Shorter, Assistant Professor, University of Michigan
Julie Rocho-Levine, Manager of Marine Animals, Dolphin Quest Oahu
Dr. Julie van der Hoop, Marie Skodowska-Curie Research Fellow, Aarhus University
Dr. Michael Moore, Senior Scientist, Woods Hole Oceanographic Institution
Prof. Kira Barton, Associate Professor, University of Michigan
  Bio-logging technology is becoming an ever more common tool for persistent monitoring of people and animals in their natural environment. As a result, the volume and type of information collected by these embedded sensing systems continues to increase, making algorithms that can accurately and efficiently classify and parameterize behavior from sensor data essential. How best to extract information from multiple sensors remains an open question. The problem becomes more challenging in cases where only sparse concurrent observations of the behavioral states are available to train and verify the accuracy of the algorithm. In this work, the authors present a systematic method to perform feature generation, feature selection and state classification from representative data collected from an example species --- bottlenose dolphins. This approach includes methods for evaluating window size selection during feature generation and the identification of specific feature sets that maximize classification performance. Additionally, the proposed framework incorporates information about state transition probabilities to further improve classification accuracy. Tag and video data for the analysis are collected from free-swimming dolphins at Dolphin Quest Oahu. The concurrent video data is scored by a human expert to create a set of observed behaviors. Results demonstrate that the algorithm is able to classify behavior with a high level of accuracy (greater than 90 percent) with 16 features and a window size of 0.6 seconds. Robustness of the proposed approach is evaluated by reducing the training data by 80 percent. The resulting classification accuracy is still above 87 percent. These results serve as the foundation for classification algorithms that can be used with data collected from animals where behavioral states can only be observed sporadically.