Current Projects

Past Projects

Fundamental Research on Intelligent and High Assurance Autonomy for Cyber-Physical Systems (CPS)

In theory, we are interested in two main areas: 

(a) intelligent situational awareness: An intelligent CPS should be able to understand its capability and limits from heterogeneous information obtained from multi-modal sensors; and 

(b) high assurance autonomy: A CPS should have a higher degree of intelligence which can cope with uncertain and varying situation with provable safety and security properties. 

To achieve these objectives, our group studies modeling, analysis and control of the stochastic hybrid system (SHS) which is a dynamical system with the interacting discrete dynamics (represents the logical/cyber behavior) and continuous dynamics (describes the physical behavior) of the CPS with uncertainties, and thus it can accurately model the CPS. Our group develops a set of SHS models (see details) and corresponding hybrid estimation (see details) and control (see details) theory and algorithms which can deal with a range of SHS computationally efficiently, and thus can be applied to various complex CPS applications. 

Cognitive-Aware Cyber-Physical Human Systems (CPHS)

Recently, we also extend our research to 

• intent inference and control for human-machine interaction and collaboration (e.g., pilot interacting with an aircraft, driver driving a car).

• cognitive autonomy for human-machine interaction and teaming.

We further extend our hybrid systems research to a wide range of problems such as 

• machine learning based-control and adaptive robust control with provable properties, which can cope with ever changing situations and uncertainties, while guaranteeing the system’s properties such as performance, safety, and/or security; and 

• consensus based state estimation for a network of hybrid-dynamical agents

• event-based state estimation for SHS (e.g., sensor/vehicle network), 

• optimal (distributed) state estimation and control for partially observable SHS (e.g., UAS control algorithm), 

• robust state estimation for SHS, and 

• fault detection and identification for state constrained SHS (e.g., cybersecurity of UAS).