Laurence J. Moore

Optimizing Transportation Problems with Multiple Objectives

Abstract Virtually all models developed for transportation problems have focused upon the optimization of a single objective criterion, namely the minimization of total transportation costs. They have generally neglected or often ignored the multiple conflicting objectives involved in the problem, the priority structure of these objectives, various environmental constraints, unique organizational values of the firm, and bureaucratic decision structures. However, in reality these are important factors which greatly influence the decision process of transportation problems. In this study the goal programming approach is utilized in order to allow for the optimization of multiple conflicting goals while permitting an explicit consideration of the existing decision environment.

ANALYSIS OF A TRANSSHIPMENT PROBLEM WITH MULTIPLE CONFLICTING OBJECTIVES

Abstract This paper presents the formulation of a goal programming model for analysis of the transshipment problem, where multiple conflicting objectives must be considered. Included are the general G. P. model for the transshipment problem, and a representative application of goal programming to such a problem. Analysis and interpretation of the G.P. solution to the problem is presented.

Analysis of a multi-criteria project crashing model

Abstract The purpose of this paper is to report on the model formulation and results of a multi-criteria time-cost (crashing) CPM project network. In the project of interest (the installation of a bank paper processing system), multiple and conflicting objectives concerning both time and cost must be incorporated into the decision analysis. The project network is, therefore, formulated as a goal programming model, including the various managerial objectives as goal constraints and associated priorities.

Workload vs scheduling policies in a dual-resource constrained job shop

Abstract The purpose of this paper is to report on research conducted to examine the effectiveness of different scheduling policies in a dual-constrained job shop under various workload conditions. The standard assumption in most job shop scheduling research has been that a 90% utilization of the shop is achieved. However, since shop utilization levels vary widely, it was hypothesized that scheduling policies that are optimum under one load condition might not be as effective under other load conditions. The model for this simulation experiment represented a job shop constrained by both labor and machines. The shop contains four machine centers with random routing of jobs through the shop. Shop workload was defined at three levels: 70, 85 and 99% utilization. Four machine scheduling rules and three labor assignment rules were tested for each of the shop workload levels, with mean job flow time as. the performance criterion. The results of the 3 × 4 × 3 factorial experiment showed that the advantage of the SPT (shortest processing time) machine scheduling rule over other rules is diminished dramatically when shop utilization is reduced from 99 to 85% or below. This same observation holds for other rules considered. The LNQ (longest queue length) labor assignment rule outperformed other rules at the 99% utilization level, but yielded no significant difference in performance at the 85% and below workload levels.

Analysis of material and capacity requirements with Q-GERT

Abstract The purpose of this paper is to report on research conducted to develop a simulation model which incorporates the basic MRP logic into the modelled production process. The objective of the modelling effort is to provide information necessary for planning and controlling of the material and capacity requirements of a production process that includes stochastic processes. The modelling and analysis procedure exploits the fact that the structure of the production process can be logically modelled as a network flow process. The technique employed to facilitate the network modelling is Q-GERT (Graphical Evaluation and Review Technique) modelling and simulation language. A case application is presented to demonstrate a prototype model, followed by a report of the simulation results. The model is designed to provide answers and insights to the following questions: (1) how long must each production centre operate to yield the specified quantities of finished products, (2) what is the expected lead time r...

Redesign of primary response areas for county ambulance services

Abstract This paper describes a modeling approach used to redesign primary response areas (sectors) for an existing county ambulance service. The model employs multicriteria analysis of the problem including goals to minimize and balance travel times and balance workloads among ambulance service units. The model is quite flexible in that optimal designs can be made subject to budgetary and personnel constraints and sector continuousness is maintained. Given inputs such as the location of the ambulance bases and demographic information, the model generates efficient and equitable designs. The model is solved using a modified multicriteria gradient search technique for nonlinear models. The sensitivity of the model to reordering planning priorities is also evaluated.

A decision support system for funds management in a public university

Senior administrators at public universities spend a great deal of time attempting to manage university funds. As the portion of total operations supported by state allocations has decreased, the number of sources of funding has been expanded by universities. This has led to a vastly more complex funds management problem for officials of public universities. The president and senior financial officer require up-to-date information on the status of university funds in terms of availability from numerous sources and applications to various uses over time. This need is more critical than ever due to legal, contractual, and legislated restrictions related to the uses of funds from the various sources. Assistance is also needed for the process of planning the allocation of funds from the various sources to potential uses over time. This paper reports on a computer based decision support system developed and implemented at a large public university to provide senior university administrators with up-to-date information on the status of funds, by source and use over time, and an interactive multicriteria model embedded within the overall computer package to assist university administrators in decision making regarding the allocation of funds from sources to uses over several time periods.

Global Analysis of a Multi-Product, Multi-Line Production System Using Q-GERT Modeling and Simulation

Abstract Numerous techniques and algorithms have been developed for the analysis of production flow line systems. However, due to the complex nature of such systems, it is often possible to analyze only a small segment of the production line system at one time. When solution designs result from this analysis and are then implemented within the overall system, inefficient results can and frequently do occur. As such, an overall methodology is needed to analyze complex production line systems in their entirety. Q-GERT network modeling and simulation offers unique possibilities for this purpose. Q-GERT, a derivative of GERT network modeling, includes such features as probabilistic branching, network looping, multiple node realizations, and queue nodes which make it applicable to complex network forms such as production flow lines. In this paper Q-GERT is demonstrated as an appropriate vehicle for analysis via a detailed case example of a multi-product, multi-line production system. The production system is f...

Combined continuous-discrete system simulation with GASP IV

Abstract GASP IV provides an alternative to the several better known and often used simulation languages, such as GPSS, SIMSCRIPT, and DYNAMO[8, 10, 11, 13, 14]. Whereas GPSS and SIMSCRIPT are discrete change languages and DYNAMO is a continuous change simulator, § GASP IV has the unique capability to perform a combination of both discrete and continuous change in the same simulation. Thus, one may perform the more traditional discrete simulation of business problems such as queueing, inventory, production, etc., or the “Forrester” (systems dynamics) type models may be simulated[2,3,4] with GASP IV. Systems that include a combination of both discrete events and continuous change variables can be readily analyzed with GASP IV.