Edward Y. Chow

Analytical redundancy and the design of robust failure detection systems

The failure detection and identification (FDI) process is viewed as consisting of two stages: residual generation and decision making. It is argued that a robust FDI system can be achieved by designing a robust residual generation process. Analytical redundancy, the basis for residual generation, is characterized in terms of a parity space. Using the concept of parity relations, residuals can be generated in a number of ways and the design of a robust residual generation process can be formulated as a minimax optimization problem. An example is included to illustrate this design methodology.

Dynamic model-based techniques for the detection of incidents on freeways

In this paper we discuss an approach to the detection of incidents on freeways. Our techniques are based on the use of a macroscopic dynamic model describing the evolution of spatial-average traffic variables (velocities, flows, and densities) over sections of the freeway. With such a model as a starting point we develop two incident detection algorithms based on the multiple model and generalized likelihood ratio techniques. We also describe a new and very simple system for processing raw data from presence-type vehicle detectors to produce estimates of the aggregate variables, which are then in turn used as the input variables to the incident detection algorithms. Simulation results using a microscopic simulation of a two-lane freeway indicate that 1) our algorithm are robust to the differences between the dynamics of actuals traffic and the aggregated dynamics used to design the detection systems; and 2) our methods appear to work as well as existing algorithms in heavy traffic conditions and work better in moderate to light traffic. Areas for future work are outlined at the end of the paper.

Bayesian Design of Decision Rules for Failure Detection

The formulation of the decision making process of a failure detection algorithm as a Bayes sequential decision problem provides a simple conceptualization of the decision rule design problem. As the optimal Bayes rule is not computable, a methodology that is based on the Bayesian approach and aimed at a reduced computational requirement is developed for designing suboptimal rules. A numerical algorithm is constructed to facilitate the design and performance evaluation of these suboptimal rules. The result of applying this design methodology to an example shows that this approach is potentially a useful one.

Status report on the generalized likelihood ratio failure detection technique, with application to the F-8 aircraft

The generalized likelihood ratio technique, a scheme for detecting and identifying abrupt changes in dynamic systems, is described in detail. Attention is given to distinguishability, detectability, and sensitivity to modeling errors. This failure detection technique is applied to a two-dimensional model of the F-8, flying at Mach .6 at 20,000 ft. in cumulus clouds.

Sequential decision rules for failure detection

Abstract : The formulation of the decision making of a failure detection process as a Bayes sequential decision problem (BSDP) provides a simple conceptualization of the decision rule design problem. As the optimal Bayes rule is not computable, a methodology that is baed on the Baysian approach and aimed at a reduced computational requirement is developed for designing suboptimal rules. A numerical algorithm is constructed to facilitate the design and performance evaluation of these suboptimal rules. The result of applying this design methodology to an example shows that this approach is a useful one. (Author)