Chung T. Nguyen

A fuzzy logic-based intelligent controller for contact management data integration

The paper is concerned with the problem of intelligent control in the selection of data, models, and processing alternatives for contact management data integration in an undersea environment. A fuzzy-logic based system, developed for automating this selection process under uncertainty, is presented. This involves the application of a network of fuzzy inference systems and a knowledge-based decision algorithm, and is implemented as a process controller. The process controller is designed to provide decision support within a hierarchical data processing framework. Two fuzzy inference systems are detailed for inferring information essential to data integration which includes system observability and contact kinematics. In addition, a novel technique for handling uncertainty in the input data to a fuzzy inference system is summarized. Experimental results, using simulated data, are presented to demonstrate the effectiveness of this concept. Recommendations for future research are also included.

The Moi Formula for Improper Exponential Definite Integrals

A general formula is developed for solving a type of improper exponential definite integral of order n in the number plane. Termed the Moi Formula, it is shown to produce substantially simpler derivations of the finite moments of a probability distribution employed for assessing stochastic randomness, such as recently published by the authors. Other applications of the integral formula are discussed.

A TEST FOR RANDOMNESS IN THREE DIMENSIONS

A mathematical method based on a nearest neighbor spatial Poisson process is described for assessing stochastic randomness in three-dimensional Euclidean space. The classical central limit theorem is invoked to obtain a normal approximation formula for testing the hypothesis of randomness. The performance of the method is evaluated with Monte Carlo simulations. A brief description is given of the software employed for implementation of the method in practice.

Performance characterization of artificial neural networks for contact tracking

Artificial neural networks (ANNs) can be exploited in a variety of information processing applications because they offer simplicity of implementation, possess inherent parallel processing characteristics and are nonlinear and less reliant on modeling of the real process. The paper is concerned with the problem of determining the performance of ANNs trained to provide estimates of contact state variables given a time series of measurements. A method is presented for determining ANN performance. Specifically, performance is shown to be intrinsically related to system observability. A performance analysis of ANNs under various observability conditions is presented along with a methodology for selecting the appropriate ANN-generated solution with a system architecture comprised of multiple clusters of ANNs.<<ETX>>

POISSON ENSEMBLES IN THREE DIMENSIONS

The authors recently derived a method for assessing stochastic randomness in three dimensional Euclidean space. The method was derived from a nearest neighbor spatial Poisson process. An alternative probability model based on a box-counting method derived from a partial sum of a Poisson series is presented in this paper. The performance of the method is evaluated through Monte Carlo simulations with synthetically constructed random distributions. A comparison between the new discrete distribution method and the initial distance method showed that a greater likelihood of detecting randomness existed among populations with the box-counting method.

Optimization process for fractal dimension estimation

An optimization process for estimating fractal dimension with the Grassberger-Procaccia correlation approach (1983) is derived in this paper. It has met with success in the estimation of dimensionality with the known attractors of Lorenz, Van Der Pol and Henon, as well as noisy time-series data. Future work will involve an attempt to employ an artificial neural network for the derived estimator.<<ETX>>