This research deals with a comprehensive study of the air traffic over the North Atlantic  (NAT) Ocean.  The main purpose of the study is to assess the fuel savings benefit of proposed changes to the separation standards in the NAT Minimum Navigation Performance Specification (MNPS) airspace.  Using the separation standards from the 1996 NAT system as the baseline, this research investigates four different separation scenarios: Reduced Vertical Separation Minima, Reduced Vertical and Longitudinal Separation Minima, Reduced Vertical and Horizontal Separation Minima, and Free Flight.  A fast time simulation model is used to investigate the effect of the separation scenarios on several measures of system performance such as fuel consumption and communication loadings.
 

Title:    Intelligent Transportation Systems

Information technology is yielding existing solutions to such transpprtation problems as congestion, safety, and pollution. ISE faculty, in collaboration with the Rutgers Center for Infrastructure and Transportation, are developing real-time monitoring and control solutions to manage traffic efficiently during disruptive occurrences such as construction activity or accidents.
 

Title:    Two-Person Zero-Sum Communication Stochastic Games

Two-person zero-sum communication stochastic games with finite state and action spaces are considered under the long-run average payoff criterion. On a restricted set of stationary strategies where every pair of actions is taken with a positive probability, a communicating stochastic game is irreducible. In order to solve a communication stochastic game by making use of the communication property, Hoffman and Karp's algorithm is applied successively over a sequence of restricted strategy spaces that gets larger until an E - optimal stationary strategy pair is obtained for any E > 0. Two equivalent algorithms are presented. Convergence issues are discussed under different ergodic structures of communicating stochastic games. Convergence of the proposed algorithm is proven for communicating stochastic games that have optimal stationary strategies with a value independent of the initial state. Examples are given to demonstrate convergence of the proposed algorithm when the specified conditions hold and a case where this algorithm does not work.
 

Title:   A Decomposition Approach for Undiscounted Two-Person Zero-Sum Stochastic Games

Two person zero-sum stochastic games are considered under the iong-run avearage expected payoff criterion. State and action spaces are assumed finite. By making of the concept of maximal communicating classes, the following decomposition algorithm is introduced for solving two-person zero-sum stochastic games: First, the state space id decomposed into maximal communicating classes. Then, these classes are organized in an hierarchical order where each level may contain more than one maximal communicating class. Best stationary strategies for the states in a maximal communication class at a level are determined by using the best stationary strategies of the states in the previous levels that are accessible from that class. At the initial level, a restricted game is defined for each closed maximal communicating class and these restricted games are solved independently. It is shown that the proposed decomposition algorithm is exact in the sense that the solution obtained from the decomposition procedure gives the best stationary strategies for the original stochastic game.
 

Title:    Substitutable Product Inventory Control

Substitutable product inventory problem is analyzed using the concepts of stochastic game theory. It is assumed that there are two substitutable products that are sold by different retailers and the demand for each product is random. Game theoretic nature of this problem is the result of substitution between products. Since retailers compete for the substitutable demand, ordering decision of each retailer depends on the ordering decision of the other retailer. Under the discounted payoff criterion, this problem is formulated as a two-person nonzero-sum stochastic game. In the case of linear ordering cost, it is shown that there exists a Nash equilibrium characterized by a pair of stationary base stock strategies for the infinite horizon problem.  This is the unique Nash equilibrium within the class of stationary base stock strategies.
 

Title:    Integrating PID Controllers into SPC Applications:  Tuning, Monitoring, and Performance Analysis

In this research, we develop a technique for integrating PID controllers into the quality control system.  We consider pure-gain and first-order processes subject to white noise disturbances and random mean shifts.  Our aim is to demonstrate that performance improvements could be achieved using off-the-shelf controllers.  We derive an expression for the sensitivity of a minimum variance (MV) control system and use this result to demonstrate that a stationary system may become extremely sensitive when mean shifts are present.  Experimental design techniques are used to demonstrate that a "sub-optimal" PID control may be implemented without producing a large negative impact on the system.  Furthermore, these results are used to develop formulas for selecting the controller parameters as well as acceptable ranges for these parameters.  The addition of PID control requires modification of the statistical process monitoring scheme.  We explore this and make recommendations for choosing the best statistic.  Cost models are introduced to assist in the evaluation of the relative advantages of each of the proposed control schemes.
 

Title:     A Generalized Forecasting Model for Predicting the Performance of Repair Stations

The objective of this research is to develop a generalized mathematical model using the modified nonhomogeneous Poisson process to (1) evaluate the performance of the repair stations, (2) identify significant covariate factors that contribute to the unfavorable rating, and (3) obtain guidelines which can be used to prioritize inspection activities. A distinguish feature of this general model is that it combines the two mutually dependent stochastic processes, one for monitoring the subsystem performance and the other for modeling the inspection visits, as the subsystem performance and the inspection visits are related.
 

Title:      Investigating Dynamic Threshold Weighted Voting Systems

In many applications, ranging from target detection to safety monitoring protection systems, a decision has to make on whether or not to accept the hypothesis based on the given information so that the probability of making a correct decision is maximized. This research focuses on the threshold weighted voting systems and develops a general methodology to evaluate the dependability of dynamic threshold weighted voting systems. A general recursive dependability time-dependent function of the systems is also investigated.
 

Title:     Designing Client-Server Type Transaction Processing Systems

This research focuses on designing and analyzing the performance of client-server systems in transaction processing environments.  Transactions bring the workload from clients and the requested services are carried out by the servers.  Our research involves detailed analysis of the underlying queueing network of clients, communication networks involved, server nodes with their disks and databases.  We pursue both analytical and simulation approaches to obtain response times and other performance measures as well as to study design issues such as work load allocation
 

Title:     Correlation Analysis of Manufacturing Systems

The purpose of this research is to study the performance effects of correlations (as proxies of temporal dependence) in manufacturing systems.  The project will identify random components of manufacturing systems with appreciable correlations; will study their impact on performance statistics; and will develop modeling technologies to capture correlations whenever present in appreciable magnitudes.  Potential random components include stochastic processes of customer demands, order lead times, machine failures, and down times.  The project will employ a novel modeling methodology, called TES/QTES, which emphasizes explicit modeling of correlations in such stochastic processes.  It will further be used to explore the impact of correlations on various manufacturing performance measures and optimal operating policies.
 

Title:     Development of Analytical Methods for Aviation Safety Risk Modeling, Assessment, and Management

This research presents a plan to develop/adapt analytical methods for aviation risk modeling, assessment, and management and to provide analytical support to the Federal Aviation Administration's Risk Management Decision Support (RMDS) Research & Development Project.  The research is decomposed into four major focal areas.  These are a) Risk Identification & Measurement, b) Risk Modeling, c) Risk Assessment, and d) Risk Management. The three-year research plan includes, among others, an investigation of multivariate statistical techniques for performance measure reduction, causal analysis of aircraft accidents using a new Aviation System Risk Model (ASRM), organizational risk profiling, uncertainty analysis, and multiobjective risk management. The models developed during the course of the proposed research will be integrated into a new proof-of-concept web-based software prototype, the Aviation Risk Analysis and Management (AvRAM) System.