Aubrey B. Poore

The classification of the dynamic behavior of continuous stirred tank reactors—influence of reactor residence time

Abstract The dynamic behavior of the continuous stirred tank reactor is analysed and classified for a variable reactor residence time. Although earlier work, treating the bifurcation to limit cycles and steady states with changing Damkohler number, yields a complete description of the problem, the evolution of multiple steady states and limit cycles is much more bizarre as the reactor residence time varies. In addition, it is the reactor residence time which is most easily varied experimentally so that the present results are more readily compared with experiment. Plots are given to show the influence of system parameters on the reactor behavior.

Multidimensional assignment formulation of data association problems arising from multitarget and multisensor tracking

The ever-increasing demand in surveillance is to produce highly accurate target and track identification and estimation in real-time, even for dense target scenarios and in regions of high track contention. The use of multiple sensors, through more varied information, has the potential to greatly enhance target identification and state estimation. For multitarget tracking, the processing of multiple scans all at once yields high track identification. However, to achieve this accurate state estimation and track identification, one must solve an NP-hard data association problem of partitioning observations into tracks and false alarms in real-time. The primary objective in this work is to formulate a general class of these data association problems as multidimensional assignment problems to which new, fast, near-optimal, Lagrangian relaxation based algorithms are applicable. The dimension of the formulated assignment problem corresponds to the number of data sets being partitioned with the constraints defining such a partition. The linear objective function is developed from Bayesian estimation and is the negative log posterior or likelihood function, so that the optimal solution yields the maximum a posteriori estimate. After formulating this general class of problems, the equivalence between solving data association problems by these multidimensional assignment problems and by the currently most popular method of multiple hypothesis tracking is established. Track initiation and track maintenance using anN-scan sliding window are then used as illustrations. Since multiple hypothesis tracking also permeates multisensor data fusion, two example classes of problems are formulated as multidimensional assignment problems.

A New Lagrangian Relaxation Based Algorithm for a Class ofMultidimensional Assignment Problems

Large classes of data association problems in multiple targettracking applications involving both multiple and single sensorsystems can be formulated as multidimensional assignment problems.These NP-hard problems are large scale and sparse with noisyobjective function values, but must be solved in“real-time”. Lagrangian relaxation methods have proven to beparticularly effective in solving these problems to the noise levelin real-time, especially for dense scenarios and for multiple scansof data from multiple sensors. This work presents a new class ofconstructive Lagrangian relaxation algorithms that circumvent some ofthe deficiencies of previous methods. The results of severalnumerical studies demonstrate the efficiency and effectiveness of thenew algorithm class.

Adaptive Gaussian Sum Filters for Space Surveillance

The representation of the uncertainty of a stochastic state by a Gaussian mixture is well-suited for nonlinear tracking problems in high dimensional data-starved environments such as space surveillance. In this paper, the framework for a Gaussian sum filter is developed emphasizing how the uncertainty can be propagated accurately over extended time periods in the absence of measurement updates. To achieve this objective, a series of metrics constructed from tensors of higher-order cumulants are proposed which assess the consistency of the uncertainty and provide a tool for implementing an adaptive Gaussian sum filter. Emphasis is also placed on the algorithms potential for parallelization which is complemented by the use of higher-order unscented filters based on efficient multidimensional Gauss-Hermite quadrature schemes. The effectiveness of the proposed Gaussian sum filter is illustrated in a case study in space surveillance involving the tracking of an object in a six-dimensional state space.

Bifurcation problems in nonlinear parametric programming

The nonlinear parametric programming problem is reformulated as a closed system of nonlinear equations so that numerical continuation and bifurcation techniques can be used to investigate the dependence of the optimal solution on the system parameters. This system, which is motivated by the Fritz John first-order necessary conditions, contains all Fritz John and all Karush-Kuhn-Tucker points as well as local minima and maxima, saddle points, feasible and nonfeasible critical points. Necessary and sufficient conditions for a singularity to occur in this system are characterized in terms of the loss of a complementarity condition, the linear dependence of the gradients of the active constraints, and the singularity of the Hessian of the Lagrangian on a tangent space. Any singularity can be placed in one of seven distinct classes depending upon which subset of these three conditions hold true at a solution. For problems with one parameter, we analyze simple and multiple bifurcation of critical points from a singularity arising from the loss of the complementarity condition, and then develop a set of conditions which guarantees the unique persistence of a minimum through this singularity.

A Numerical Study of Some Data Association Problems Arising in Multitarget Tracking

The central problem in multitarget/multisensor tracking is the data association problem of partitioning the observations into tracks and false alarms so that an accurate estimate of the true tracks can be recovered. This data association problem is formulated in this work as a multidimensional assignment problem. These NP-hard data association problems are large scale, have noisy objective functions, and must be solved in real-time. A class of Lagrangian relaxation algorithms has been developed to construct near-optimal solutions in real-time, and thus the purpose of this work is to demonstrate many of the salient features of tracking problems by using these algorithms to numerically investigate constant acceleration models observed by a radar in two dimensional space. This formulation includes gating, clustering, and optimization problems associated with filtering. Extensive numerical simulations are used to demonstrate the effectiveness and robustness of a class of Lagrangian relaxation algorithms for the solution of these problems to the noise level in the problems.

Numerical Continuation and Singularity Detection Methods for Parametric Nonlinear Programming

Numerical methods are developed for continuation, solution-type determination, and singularity detection in the parametric nonlinear programming problem. This problem is first converted to a closed, “active set” system of equations

A bifurcation analysis of the nonlinear parametric programming problem

\bar F ( z,\alpha ) = 0

Variable order Adams-Bashforth predictors with an error-stepsize control for continuation methods

, which includes a nonstandard normalization of the multipliers. A framework is then developed for combining various numerical continuation methods with a large number of null and range space methods from constrained optimization. By exploiting the special structure in the parametric optimization problem, solution-type classification and singularity detection are shown to require minimal additional expense beyond that involved in the continuation procedure itself. Due to the special structure of these problems, singularity detection methods are more comprehensive than those for general nonlinear equations. In this development, the Schur complement and related results play an important and unifying role. As an illustration, these methods are used to prod...

New multidimensional data association algorithm for multisensor-multitarget tracking

The structure of solutions to the nonlinear parametric programming problem with a one dimensional parameter is analyzed in terms of the bifurcation behavior of the curves of critical points and the persistence of minima along these curves. Changes in the structure of the solution occur at singularities of a nonlinear system of equations motivated by the Fritz John first-order necessary conditions. It has been shown that these singularities may be completely partitioned into seven distinct classes based upon the violation of one or more of the following: a complementarity condition, a constraint qualification, and the nonsingularity of the Hessian of the Lagrangian on a tangent space. To apply classical bifurcation techniques to these singularities, a further subdivision of each case is necessary. The structure of curves of critical points near singularities of lowest (zero) codimension within each case is analyzed, as well as the persistence of minima along curves emanating from these singularities. Bifurcation behavior is also investigated or discussed for many of the subcases giving rise to a codimension one singularity.