J. Hugh Ellis

Assessing the impact of airline travel on the geographic spread of pandemic influenza.

The objective of this research is to explore what would happen if the Hong Kong influenza pandemic strain of 1968–1969 returned in 2000. We report the results of a series of simulations of an SEIR epidemic model coupled with air transportation data for 52 global cities. Preliminary results suggest that if the 1968–1969 pandemic strain returned, it would spread concurrently to cities in both the northern and southern hemispheres thereby exhibiting less of the characteristic seasonal swing. In addition, after recognition of pandemic onset in the focal city, the time lag for public health intervention is very short. These findings highlight the importance of coordinated global surveillance and pandemic planning.The objective of this research is to explore what would happen if the Hong Kong influenza pandemic strain of 1968–1969 returned in 2000. We report the results of a series of simulations of an SEIR epidemic model coupled with air transportation data for 52 global cities. Preliminary results suggest that if the 1968–1969 pandemic strain returned, it would spread concurrently to cities in both the northern and southern hemispheres thereby exhibiting less of the characteristic seasonal swing. In addition, after recognition of pandemic onset in the focal city, the time lag for public health intervention is very short. These findings highlight the importance of coordinated global surveillance and pandemic planning.

Multiobjective air pollution monitoring network design

Abstract The methodologies currently available to optimize the locations of air pollutant monitoring stations typically include a single pollutant and a single objective. In this paper, multiple objective functions are introduced that provide performance measures describing the spatial coverage of the network and its ability to detect violations of standards for multiple pollutants. An additional objective regarding the effect of data validity in the design is also considered. The objective functions are cast in a manner that permits considerable flexibility in the model formulation. The numerical difficulties associated with the formulation are discussed as are extensions of the basic framework. The basic model and its extensions have been applied to the design of a monitoring network for Tarragona, Spain.

Modeling of risk-based inspection, maintenance and life-cycle cost with partially observable Markov decision processes

The utilization of Markov decision processes as a sequential decision algorithm in the management actions of infrastructure (inspection, maintenance and repair) is discussed. The realistic issue of partial information from inspection is described, and the classic approach of partially observable Markov decision processes is then introduced. The use of this approach to determine optimal inspection strategies is described, as well as the role of deterioration and maintenance for steel structures. Discrete structural shapes and maintenance actions provide a tractable approach. In-service inspection incorporates Bayesian updating and leads to optimal operation and initial design. Finally, the concept of management policy is described with strategy vectors.

STOCHASTIC PROGRAMS FOR IDENTIFYING CRITICAL STRUCTURAL COLLAPSE MECHANISMS

Abstract An important task in structural analysis is the identification of critical structural collapse mechanisms, that is, given a structural design and loading configuration, determine which failure mechanisms will occur first. In this work we extend traditional deterministic optimization models for failure mode identification to stochastic forms by considering external loads and structural member plastic moment capacities as correlated random variables. A hybrid model is first developed that contains both chance constrained programming and stochastic linear programming features. A purely chance constrained model is then described. Both models represent nonconvex programming problems. Computational experience with the models is described through an application to the analysis of a portal frame.

An Analytical Solution to the One-Dimensional Solute Advection-Dispersion Equation in Multi-Layer Porous Media

An analytical solution to the one-dimensional solute advection-dispersion equation in multi-layer porous media is derived using a generalized integral transform method. The solution was derived under conditions of steady-state flow and arbitrary initial and inlet boundary conditions. The results obtained by this solution agree well with the results obtained by numerically inverting Laplace transform-generated solutions previously published in the literature. The analytical solution presented in this paper provides more flexibility with regard to the inlet conditions. The numerical evaluation of eigenvalues and matrix exponentials required in this solution technique can be accurately and efficiently computed using the sign-count method and eigenvalue evaluation methods commonly available. The illustrative calculations presented herein have shown how an analytical solution can provide insight into contaminant distribution and breakthrough in transport through well defined layered column systems. We also note that the method described here is readily adaptable to two and three-dimensional transport problems.

On the equivalence of kriging and maximum entropy estimators

This study compares kriging and maximum entropy estimators for spatial estimation and monitoring network design. For second-order stationary random fields (a subset of Gaussian fields) the estimators and their associated interpolation error variances are identical. Simple lognormal kriging differs from the lognormal maximum entropy estimator, however, in both mathematical formulation and estimation error variances. Two numerical examples are described that compare the two estimators. Simple lognormal kriging yields systematically higher estimates and smoother interpolation surfaces compared to those produced by the lognormal maximum entropy estimator. The second empirical comparison applies kriging and entropy-based models to the problem of optimizing groundwater monitoring network design, using six alternative objective functions. The maximum entropy-based sampling design approach is shown to be the more computationally efficient of the two.

Reliability-based bridge design and life cycle management with Markov decision processes☆

Abstract A Markov Decision Process-based (MDP) modelling approach is used to identify optimal structural designs and their associated maintenance policies. This approach has the desirable attributes of capturing the dynamics of the coupled structural design/maintenance problem while yieldding a mathematical model that is computationally tractable. Implicit in the MDP is the computationally and theoretically usful choice of modelling the dynamics of structural performance and maintenance actions as two related but separate processes. Structural resistance deterioration due to corrosion and stochastic structural loading are modelled in the MDP as a “self-transition” process, that is, a process that models structural performance without maintenance intervention. A suite of maintenance actions and their associated costs constitute the core of the “decision effect” process. These two processes, made computationally operational through the development of self-transition and decision effect matrices, are then combined into a joint transition matrix, with which we can model the effects of maintenance actions on bridge performance and costs. Initial designs and their associated minimum expected, discounted lifetime costs are identified with the MDP for a two girder bridge.

Multiobjective mathematical programming models for acid rain control

Abstract Although acid rain control is inherently multiobjective, previous optimization approaches have generally been single-objective, often acting to minimize aggregate abatement cost or emission reductions. Using an updated, least-cost deposition-constrained deterministic model as a basic framework, three multiobjective models are developed that consist of formulations which permit deviations about target deposition levels, the addition of constraints to effect measures of equity and models to enforce restrictions on aggregate emission reduction tonnage. The deposition deviation model shows that large abatement cost savings can be realized if the hard upper bound on maximum allowable deposition limit is preferentially relaxed. The socalled equity model develops strategies that attempt to balance within each state and province, the disparity between fractional emission and fractional deposition reductions. The aggregate emission reduction model shows some of the effects associated with the imposition of a common type of acid rain proposal. Our intent is to demonstrate that the incorporation of multiobjectivity into mathematical programming models for optimizing acid rain control constitutes an important step toward the identification of more representative, more useful and hopefully, scientifically and politically acceptable abatement strategies.

Integrating multiple long-range transport models into optimization methodologies for acid rain policy analysis

Abstract This paper addresses certain complications inherent in developing efficient acid rain control strategies given that our knowledge of the physics and chemistry of long-range pollutant transport and transformation is imprecise. This imprecision (or uncertainty) translates, in a sense, into risk. We seek here to demonstrate that there are currently available methods to better define this risk and provide a basis for decision-making. More specifically, these methods involve chance-constrained stochastic programming techniques. A distinguishing characteristic of the approach is our use of the output from multiple long-range transport models to develop probabilistic descriptors of transport/transformation phenomena.

System-identification procedure for system and input parameters in ambient vibration surveys

Abstract Over the past several decades, much valuable data pertaining to the structural performance of large civil structures (e.g., bridges) have been obtained through the use of ambient vibration surveys. While there have been significant advances in the technology used to collect the data, processing techniques used in such situations are still unable to reliably extract “difficult” parameters such as damping. This paper presents an overview of a method based on the application of a Kalman filter for the identification of system and excitation characteristics. Numerical examples of the method are presented.