My research interest is in hybrid system modeling, hybrid estimation algorithms and applications.

A hybrid system contains interacting continuous state and discrete state. The continuous state dynamics is usually described by a set of differential equations (or difference equations). In many stochastic hybrid system models, the discrete state evolution (or mode transition) is assumed to be a Markov process with a constant mode transition probability independent of the continuous state evolution. However, in many hybrid systems, the mode transitions may be induced by the evolution of the continuous state. An example is an aircraft subsystem that changes from an `operational' state to a `faulty' state when its temperature goes beyond a critical value Tc. I am interested in hybrid systems that have discrete state transitions (or mode transitions) that depend stochastically on the continuous state. These kinds of discrete state transitions could be described mathematically by stochastic guards. In [4, 6], I have proposed a Stochastic Linear Hybrid System model for hybrid system with discrete state transitions governed by stochastic guards.

I have investigated efficient algorithms to solve the hybrid estimation problem based on Gaussian mixture approximations. In [4, 6]. I have proposed hybrid estimation algorithm for the Stochastic Linear Hybrid System. In [1, 3], I proposed a hybrid estimation algorithm based on a continuous-state dependent mode transition matrix and a Monte Carlo integration method. The convergence property of the Monte Carlo integration is investigated in [3]. In [2, 6], I proposed a hybrid estimation algorithm based on based on a continuous-state dependent mode transition matrix and an analytical integration method.

One application of the hybrid system model/algorithm is in aircraft tracking in Air Traffic Control [5]. I am interested in other possible applications in the areas such as fault diagnosis, aircraft trajectory prediction and intent inference.

I am currently working on the analysis of hybrid estimation algorithms that are made up of a bank of interacting Kalman filters. An online recursive algorithm to predict the performance of such algorithms has been proposed in [7].

Publications:

[1] C.E. Seah and I. Hwang, Hybrid Estimation Algorithm using State-Dependent Mode Transition Matrix for Aircraft Tracking, submitted to AIAA Guidance, Navigation, and Control Conference, August 2006.

[2] Hwang and C.E. Seah. An estimation algorithm for stochastic linear hybrid systems with continuous-state-dependent mode transitions. In Proceedings of the 45th IEEE Conference on Decision and Control, pp. 131-136, San Deigo, CA, USA, December 2006.

[3] C.E. Seah and I. Hwang. Target tracking of arrival aircraft using hybrid estimation. In 44th Annual Allerton Conference on Communication, Control, and Computing, Monticello, IL, September 2006.

[4] C.E. Seah and I. Hwang. Hybrid estimation for stochastic piecewise linear systems. In Hybrid Systems: Computation and Control, Lecture Notes in Computer Science, Springer-Verlag, 2007.

[5] C.E. Seah and I. Hwang. A hybrid estimation algorithm for terminal-area aircraft tracking. Submitted to AIAA Guidance, Navigation and Control Conference, 2007, accepted.

[6] C.E. Seah and Inseok Hwang. Stochastic Linear Hybrid Systems: Modeling, Estimation, and Applicationin Air Traffic Control. Submitted to IEEE Transactions on Control Systems Technology, 2007, under revision.

[7] C.E. Seah and Inseok Hwang. Performance Analysis of Hybrid Estimation Algorithms by Characterization of Kalman Filter Residuals. Submitted to 46th IEEE Conference on Decision and Control, 2007.