Gang Tao received his B.S. (EE) degree from University of Science and Technology of China in 1982, M.S. (EE, CpE and APMA) degrees and Ph.D. (EE) degree from University of Southern California during 1984-1989. He worked in the areas of adaptive control, with particular interests in adaptive control of systems with multiple inputs and multiple outputs and with nonsmooth nonlinearities and actuator failures, in stability and robustness of adaptive control systems, and in passivity characterizations of control systems. His publications include the 2003 Wiley textbook "Adaptive Control Design and Analysis" (by Gang Tao), the 1996 Wiley book "Adaptive Control of Systems with Actuator and Sensor Nonlinearities" (by Gang Tao and Petar Kokotovic), the 2004 Springer book "Adaptive Control of Systems with Actuator Failures" (by Gang Tao, Shuhao Chen, Xidong Tang and Suresh M. Joshi), the 2003 Springer book "Control of Sandwich Nonlinear Systems" (by Avinash Taware and Gang Tao), the 2001 Springer book "Adaptive Control of Nonsmooth Dynamic Systems" (edited by Gang Tao and Frank Lewis), the 2009 book "Advances in Control Systems Theory and Applications" (edited by Gang Tao and Jing Sun), and over 320 technical papers and book chapters. Recently he has been working on adaptive control of systems with uncertain actuator failures and actuator nonlinearities, with structural damage and sensor uncertainties and failures, for aircraft and spacecraft flight control applications.
He is an associate editor for Automatica and a subject editor for International Journal of Adaptive Control and Signal Processing (for which he was a guest editor for a 1997 special issue on Adaptive Systems with Nonsmooth Nonlinearities). He was an associate editor for IEEE Trans. on Automatic Control from 1996 to 1999. He organized and chaired the 2001 International Symposium on Adaptive and Intelligent Systems and Control, held in Charlottesville, Virginia, and organized invited sessions for 1996 IEEE CDC and 1999 IEEE CCA on adaptive control of systems with nonsmooth nonlinearities. He served on numerous international conferences' technical and advisory committees and as Technical Conference Chair for IEEE SoutheastCon 2007.
He is a Fellow of IEEE.
Adaptive Fault Accommodation Based Resilient Control Systems
Resilience, literally, is the ability to recover readily from illness, depression or adversity. For control systems, illness, depression or adversity can be faults, disturbances or uncertainties, so one may define resilience as the system's ability to accommodate its faults, disturbances and uncertainties. The goal of resilient control is to increase the control system resilience for performance guarantees under such adverse circumstances. Recently, the developments of new concepts, theory and techniques of resilient control systems have been an important research focus of major practical urgency.
System faults such as component damage, actuator and sensor failures, can cause large system structural and parametric uncertainties which may impose challenges to existing feedback control design methods. For performance-critical and emerging control system applications (such as aircraft and spacecraft flight control, MEM actuators and sensor networks, cyber physical systems, multi-agents, smart grids), a key demand on fault accommodation is that, in addition to stability, certain desired system performance such as tracking or optimality is still ensured despite system faults. Thus, a resilient control system is required to have the capacity to maintain desired performance in the presence of uncertain and multiple faults. Some existing control methods cannot meet such a requirement. New resilient control theory and techniques need to be developed to meet the new control demands.
In this talk, we demonstrate how adaptive control techniques are used to increase system resilience to overcome the effects of uncertain actuator failures and structural damage, by adaptively compensating such failures and damage, to meet the desired system performance. We also show how adaptive control designs are expanded to accommodate multiple uncertain system faults, by using an adaptive multi-design integration based resilient control strategy.
To deal with unknown structural variations caused by multiple faults, multiple controllers may have to be used. A multiple-model controller works well if it is properly parametrized and its control switching mechanism is properly designed, and adaptation is needed if there are system uncertainties. To deal with multiple faults, multi-structure controller parametrizations are used, each corresponding to a fault pattern, and to improve system performance, a multiple-group based multiple-model control method is employed. To ensure desired system performance, adaptive multi-design integration based control schemes can be used, as effective designs for resilient control systems.
We will present several new adaptive multiple-model control schemes and adaptive multi-design integration control schemes, which are capable of handling different types of faults and uncertainties, by using either a single reference model system, a set of reference model systems, or a set of multi-structure controllers, to achieve desired closed-loop system parametrization and adaptation. We will show the desired performance of such schemes when applied to aircraft flight control system models. We will also discuss some related technical issues in such an adaptive framework for resilient control systems.