Dr. Hari Krishna Garg

A Team from NUS, comprising Associate Professor Hari Krishna Garg, students Kartik Prabhakara and Brojo Pillai, won the Bronze medal at SIMagine 2002.

"We are very excited to have been able to showcase NUS at such a prestigious global event. With support from the University in entrepreneurial activites, and the support and encouragement from our ECE Department, we are confident that we will go a long way." commented Associate Professor Hari Krishna Garg, on receiving the award.

The team received a trophy and a cash prize of Euro 10,000 (S$16310) in Cannes, France. The prestigious worldwide JavaCard application development competition is organized by SchlumbergerSEMA, SUN Microsystems, Motorola and numerous European mobile operators.

The team from NUS came 1st among the University entries and placed 3rd overall, competing with nearly 200 entries from established campanies, universities and independent developers around the world.

The NUS team has formed a start-up company, Purple Ace, and is looking into refining and commercializing its product.

------------------------------------------------------
For further information, please contact:
Dr. Hari Krishna Garg
Email: eleghk@nus.edu.sg
Telephone: 6516 4542

Dr. Ge Shuzhi, Sam

Presented at the 4th World Congress on Intelligent Control and Automation (WCICA), Shanghai, China, 10-14 June 2002. Professor Ge has made outstanding contributions to the fundamental theory underpinning intelligent control systems, and to practical engineering methodologies. In great contrast to the common empirical approaches, he did extensive research in the field of stable on-line adaptive neural control. Stability of the closed-loop systems is rigorously proven through Lyapunov stability analysis.

Owing to fundamental contributions made, stable adaptive neural control can be constructed not only for typical benchmark nonlinear systems such as affine nonlinear systems, but also for more general unknown complex systems that are beyond the scope of traditional model-based control approaches, such as nonaffine nonlinear systems, or pure-feedback non-linear systems, with mathematical rigor. It clearly demonstrates that neural control is the next step of development in adaptive control systems and intelligent control, and beyond, as conventional model-based control is of little use for these complex problems.

Robust Adaptive Neural Control for a Class of Perturbed Strict Feedback Nonlinear Systems

Abstract >> This paper presents a robust adaptive neural control design for a class of perturbed strict feedback nonlinear system with both completely unknown virtual control coefficients and unknown nonlinearities. The unknown nonlinearities comprise two types of nonlinear functions: one naturally satisfies the "triangularity condition" and can be approximated by linearly parameterized neural networks, while the other is assumed to be partially known and consists of parametric uncertainties and known "bounding functions." With the utilization of iterative Lyapunov design and neural networks, the proposed design procedure expands the class of nonlinear systems for which robust adaptive control approaches have been studied.

The design method does not require a priori knowledge of the signs of the unknown virtual control coefficients. Leakage terms are incorporated into the adaptive laws to prevent parameter drifts due to the inherent neural-network approximation errors. It is proved that the proposed robust adaptive scheme can guarantee the uniform ultimate boundedness of the closed-loop system signals. The control performance can be guaranteed by an appropriate choice of the design parameters. Simulation studies are included to illustrate the effectiveness of the proposed approach.

IEEE Transactions on Neural Networks
Publication Date: Nov 2002
Volume: 13, Issue 6 on page(s): 1409-1419

------------------------------------------------------
For further information, please contact:
Dr. Ge Shuzhi, Sam
Email: elegesz@nus.edu.sg
Telephone: 6516 6821