Research

Projects

• On-going Projects

- Communication network design for autonomous vehicle platoon in smart city environment, National Research Foundation of Korea, 2025.03.01 ~ 2030.02.28

• Past Projects

- Autonomous vehicle platoon in smart city environment, Sookmyung Women's University Research Grants, 2024.03.01 ~ 2025.02.28

- Research on cybersecurity and resilient control for autonomous vehicles, Sookmyung Women's University Research Grants, 2023.03.01 ~ 2024.02.29

- Development of sensorless force control algorithm based on passivation approach, Sookmyung Women's University Research Grants, 2022.03.01 ~ 2023.02.28

- Development of pose-graph optimization and node simplification algorithm for multi-agent systems, LG Electronics, 2022.05.01 ~ 2022.12.31

Research Interests

• Platooning of autonomous vehicles

With the interest in autonomous vehicles, lots of research on the platooning of autonomous vehicles has been conducted to improve driver convenience and reduce the risk of accidents. Furthermore, by reducing the distance between vehicles, energy efficiency can be increased due to air resistance. However, when the distance is reduced, a string stability problem may occur, in which small changes in the speed of the leading vehicle are amplified and transmitted to the following vehicles. Thus, in this research, we focus on the stability of formation control for autonomous vehicles that exchange information with urban infrastructure and surrounding vehicles using V2X(vehicles-to-everything) communication.

• Safe human-robot interaction

With increasing demands on human-robot collaboration in industrial, commercial, and home environments, collaborative robots have attracted much attention. Compared to industrial robots requiring high stiffness and position accuracy for operating structured environments, collaborative robots need to adapt to unstructured environments and perform safe operations between the robot and humans or objects to avoid unintended damage to the environment. In this research, we focus on the force control algorithm of collaborative robots for implementing human-robot safe cooperation in a shared workspace.

• Cooperative control of multi-agent systems

Over the last decade, cooperative control problems of multi-agent systems have attracted remarkable attention due to various applications such as cooperative control of unmanned aerial vehicles, communication among sensor networks, and control and optimization of distributed generation and storage devices in smart grids. 

• Cybersecurity on cyber-physical systems

With the progress of computing and communication technologies, diverse IT infrastructures across the different system layers and heterogeneous physical plants are integrated to improve efficiency. Such a cyber-physical framework, called as a cyber-physical system (CPS), has been applied to many industrial fields, including the smart grids, healthcare systems, and autonomous vehicles. However, by increasing the complexity of the network, the vulnerability of CPS from malicious attackers is escalated. For example, it has been reported that such attacks caused severe damage to critical infrastructure. The Natanz uranium enrichment facility in Iran was infected by Stuxnet malware. In 2015, cyber attacks on the Ukraine power grid compromised three regional distribution companies and it led to a 6-h blackout. Thus, in this research, we deal with an attack-resilient control structure for a CPS to enhance the CPS security against stealthy system integrity attacks that manipulate the state of the physical plant while undetected.

• Simultaneous localization and mapping (SLAM)

To deal with many tasks involving guide, surveillance, transportation, and automated vacuum cleaning, building a map of the environment and localizing its current pose in the obtained map is an essential ability for mobile robots. One way to formulate simultaneous localization and mapping (SLAM) problems is to use a graph whose nodes correspond to the robot poses and whose edges represent constraints between the poses. However, for the long-term operation of mobile robots, there are analogous measurement data providing similar information since a robot repeatedly navigates the same place. The redundant edges should be removed for resource-constrained pose-graph SLAM problems. In this research, we present node marginalization and edge sparsification algorithms based on the graph structure to the size of data while minimizing information loss.

• Disturbance estimation and compensation

The primary objective of control may be to make a system response satisfy a given specification such as the overshoot, settling time, steady-state error between the reference input and system output, and so on. When there is no modeling error, it is easy to achieve the given specification by a simple unity feedback control system. However, it is impossible to obtain a precise mathematical model from the actual plant because there are some limitations to obtaining exact information about the actual plant. Moreover, in real control situations, the existence of disturbance and measurement noise is also inevitable. Thus, in this research, we focus on a design of disturbance estimation and compensation controller as a simple and powerful tool for robust control to reject the disturbances and model uncertainties.

• Nonlinear system analysis

The concept of passivity has become a popular research subject after its introduction in the control community. From the viewpoint of energy dissipation, it allows us to characterize a complex dynamical system in terms of a simple input–output relation, which enables simple analysis and control synthesis. To be specific, a passive system connected through a negative feedback to another strictly passive system (e.g., static gain) is asymptotically stable regardless of their internal dynamics and nonlinearities. Another notable virtue of passivity is that a parallel or a negative feedback interconnection of passive systems are also passive. These properties are powerful tools in analyzing interconnected systems and/or synthesizing controls. Thus, we will further investigate the detailed properties of passive systems to understand and exploit complicated nonlinear systems.