Data-Driven Adaptive Control
Retrospective Cost Adaptive Control RCAC was extended by the inclusion of RLS-based online identification. DDRCAC (Data-Driven RCAC) further reduces the need for modeling information compared to RCAC. Furthermore, DDRCAC can be applied to MIMO systems with uncertain unstable zero dynamics. Important applications of DDRCAC include hypersonic test vehicle (HTV-2), planar interceptor control, flutter suppression, and adaptive failure recovery.

Funded by Office of Naval Research.

Predictive Cost Adaptive Control
A novel combination of model-predictive and adaptive control was developed in collaboration with Tam W. Nguyen and Ilya V. Kolmanovsky. This method can accommodate constraints and utilizes output-feedback models, while using minimal modeling information.

Funded by Office of Naval Research.

Adaptive Control and Its Applications
Research was done on Retrospective Cost Adaptive Control (RCAC) and its applications in adaptive PID, decentralized, and MIMO control in collaboration with Yousaf Rahman, Antai Xie, Ankit Goel, Adam L. Bruce, Juan Augusto Paredes, Nima Mohseni, and Mohammadreza Kamaldar. A particular focus of research was the control of NMP plants and understanding the mechanisms underlying RCAC.

Funded by Office of Naval Research, Air Force Office of Scientific Research, and National Science Foundation.

Fundamentals of Digital Algorithms
Real-time implementation of digital systems is an intricate subject that requires careful attention. To this end, fundamental issues of real-time digital implementations of Recursive Least Squares and Kalman Predictor and Filter, were explored in a series of publications.

Funded by Air Force Office of Scientific Research.

Phasor-Based Control for Scalable Solar Photovoltaic Integration
This project to developed and tested new methods of grid control using phasor measurements, in the presence of significant grid-connected photovoltaic generation. This research involved the application of adaptive control for regulating an electric grid model using little or no prior modeling information.

Funded by United States Department of Energy.