Jeffery K. Cochran

A multi-population genetic algorithm to solve multi-objective scheduling problems for parallel machines

In this paper we propose a two-stage multi-population genetic algorithm (MPGA) to solve parallel machine scheduling problems with multiple objectives. In the first stage, multiple objectives are combined via the multiplication of the relative measure of each objective. Solutions of the first stage are arranged into several sub-populations, which become the initial populations of the second stage. Each sub-population then evolves separately while an elitist strategy preserves the best individuals of each objective and the best individual of the combined objective. This approach is applied in parallel machine scheduling problems with two objectives: makespan and total weighted tardiness (TWT). The MPGA is compared with a benchmark method, the multi-objective genetic algorithm (MOGA), and shows better results for all of the objectives over a wide range of problems. The MPGA is extended to scheduling problems with three objectives: makespan, TWT, and total weighted completion times (TWC), and also performs better than MOGA.

Statistically Constrained Economic Design of the EWMA Control Chart

This paper presents a statistically constrained economic model for the optimal design of an exponentially weighted moving average (EWMA) control chart for controlling process means. The optimum design parameters include the sample size, control limit wi..

Fuzzy multi-criteria selection of object-oriented simulation software for production system analysis

Discrete event computer simulation is one of the most widely used modeling tools for production systems. Object-oriented programming is revolutionizing computer simulation software packages. A fuzzy set approach for multi-criteria selection of object-oriented simulation software for analysis of production system is developed in this paper. The approach uses fuzzy set theory and algebraic operations of fuzzy numbers to characterize simulation software so that the strength and weakness of each alternative can be compared. The linguistic input from experts and decision makers are fuzzified and the results of the fuzzy inference are translated back to linguistic explanation. Further, by aggregating decision makers preference (weighting) to the experts rating, a single number for each candidate software can be obtained. One can use this number as an index to choose the most suitable simulation software for the specific application. A comparison between evaluations using simple triangular fuzzy numbers and using the real fuzzy set are presented.

A multi-stage stochastic methodology for whole hospital bed planning under peak loading

We propose a multi-stage stochastic methodology to balance inpatient bed unit utilisations in an entire hospital. It minimises blocking of beds from upstream units, within given constraints on bed reallocation, while considering multiple patient types. Queuing network analysis and optimisation are used to achieve balanced targets of bed unit utilisation while building hospital staff involvement. Discrete event simulation is then used to maximise flow through the system including nonhomogeneous effects of daily and hourly peak loading, nonexponential lengths of stay, and blocking behaviour. A 400 plus bed major hospital is analysed with the methodology and results are validated against field data.

Metamodels of production line transient behaviour for sudden machine breakdowns

SUMMARY Production controls based on mean value analysis and steady-state conditions are incapable of making on-time decisions that cope with unexpected dynamic events which are due to interruptions in complex work flow characteristic of modern production lines. One of the most significant dynamic events in production situations is a sudden machine/operator breakdown/outage. Computer simulation is known to be a useful tool for modelling the dynamic response of a system to machine breakdown, but it takes too long to perform its analysis to provide the response time necessary for control procedures. Analytical methods provide formulae which are suitable for real-time analysis, but there are no results available for transient behaviour associated with machine breakdown. Metamodelling is the process of summarizing the results of a simulation study in analytical form. In this paper, we present our metamodels for the dynamic behaviour of both time in system and number in system for a general arrival time, gener...

Optimal design of a hybrid push/pull serial manufacturing system with multiple part types

The objective of this research is to investigate the possibility of integrating Manufacturing Resource Planning (MRP)/push and Just in Time (JIT)/pull strategies in a multiproduct multistage serial manufacturing system. Each workstation is able to undertake different operations and so to produce more than one type of in-process item. A modified version of a horizontally integrated hybrid push/pull production system is developed. The system can be optimized by locating points of integration, and determining the optimal values of safety stocks for the push part and numbers of Kanbans for the pull part. A modification of genetic algorithm (GA) chromosomes and crossover procedures is developed for the optimization. The optimization involves evaluations of stochastic performance measures by a discrete event simulation model. The motivating case study of a tube shop in an aerospace manufacturer is presented. This research extensively explores the question of whether each part type should have its own junction point (less constrained) or whether there should be one common junction point for the overall system (easier to implement).

Ascertaining Important Features For Industrial Simulation Environments

Recent years have witnessed the development and commercial release of multiple simulation tools, environments, and intelligent simulators. Each release seems to contain additional advanced features designed to simplify simulation use and increase the productivity of model builders. But to date, no one has addressed feature definition from the viewpoint of a simulation practitioner. This paper discusses our efforts to identify and prioritize simulation features deemed most desirable from the practitioner viewpoint. A series of three questionnaires was developed and administered to a group of qualified simulation practitioners. With results that are of interest to simulation users, researchers, and simulation software developers, the survey responses reveal not only what practitioners feel are the most important features of presently available commercial packages, but also identify important areas for future development.

Two-stage simulation optimization for agile manufacturing capacity planning

Capacity planning involves the selection of manufacturing technologies and the allocation of budget to specific equipment acquisitions. In todays highly volatile manufacturing world, an agile capacity-planning tool is required. This tool must provide the mechanism for a company to thrive in an environment of uncertainty. Uncertain future demands make capacity planning and technology selection difficult tasks, whether they are caused by variations in forecasts of direct demand or by upstream variability in a supply chain. In this paper, a practical modelling technique for minimizing the required investment in capacity planning for discrete manufacturing sites under an uncertain demand stream is presented. The method consists of a two-stage stochastic integer program. The first stage characterizes the optimal response of the system under uncertainty. The second stage selects a tool set based on the characterization from the first stage, with the addition of budget constraints. The model is scalable, allowing for multiple products, multiple operations, multiple flow paths including re-entrant flow, and multiple tool types. A simple example is introduced to explain the methodology, followed by the results of a large-scale real-world application in the semiconductor industry.

A queuing-based decision support methodology to estimate hospital inpatient bed demand

Hospital inpatient bed capacity might be better described as evolved than planned. At least two challenges lead to this behaviour: different views of patient demand implied by different data sets in a hospital and limited use of scientific methods for capacity estimation. In this paper, we statistically examine four distinct hospital inpatient data sets for internal consistency and potential usefulness for estimating true patient bed demand. We conclude that posterior financial data, billing data, rather than the census data commonly relied upon, yields true hospital bed demand. Subsequently, a capacity planning tool, based upon queuing theory and financial data only, is developed. The delivery mechanism is an Excel spreadsheet. One adjusts input parameters including patient volume and mix and instantaneously monitors the effect on bed needs across multiple levels of care. A case study from a major hospital in Phoenix, Arizona, USA is used throughout to demonstrate the methodologies.

Optimization of a complex flow line for printed circuit board fabrication by computer simulation

Abstract The complex operations and considerable process time variability of printed circuit board (PCB) fabrication create difficulties in finding effective and efficient planning techniques for todays PCB production management. A great deal of money is involved. By modeling and testing a real world PCB fabrication facility, this paper shows that computer simulation can provide a viable planning tool to estimate production capacity and to explore optimum arrangement in batch work size of key bottleneck machines to minimize product throughput time. Many simulation experiments are performed and the results analyzed as a response surface. The general characteristic of product throughput time is found to be that its minimal value exists when batch job numbers of subsequential key machines are matched in batch size or in multiples thereof. A nonlinear empirical equation to estimate product throughput time has been derived from the simulation results.