Search

This paper investigates the applicability of particle swarm optimisation (PSO) to the dynamic lot sizing problem with batch ordering. Backorders are allowed to account for discrepancies created by the batch ordering constraint. The PSO solution is compared with solutions generated using both a modified Silver-Meal (MSM) heuristic and a genetic algorithm (GA). A 23 factorial experiment is used to compare the various approaches and to examine the influence of three factors on their performance. The investigated factors include the demand pattern, the batch size, and the planning horizon. The comparisons are based on the relative frequency of obtaining the optimum solution and the percentage deviation from the optimum solution. In general, the PSO outperformed both the MSM and the GA by producing the lowest cost solution on almost all experimental runs. The planning horizon is the most significant factor affecting the performance of all heuristics. The MSM is affected by all investigated factors while the PS...

In this paper, genetic algorithms are applied to the scheduling of a manufacturing system that is designed to support an assembly-driven differentiation strategy in the context of agile manufacturing. The system consists of a single flexible machine followed by multiple identical assembly stations. The objective of the scheduling problem is to minimize the makespan. A modified version of the genetic algorithm, inspired by the particle swarm optimization approach, is applied to the problem in addition to the general application of genetic algorithms. The objective is to investigate the potential that the particle swarm optimization concepts may have in improving the performance of genetic algorithms when applied to the chosen problem. The performance of these algorithms is compared to existing heuristics in the literature. A 2^3 factorial experiment, replicated twice, is used to compare the performance of the various approaches and identify the significant factors that affect the average percentage deviation from a lower bound. The results show that both versions of genetic algorithms applications outperform the existing heuristics in many instances and provide schedules that are shorter by as much as 15.5% in the cases considered. In addition, the modified application of genetic algorithms outperforms the regular application with shorter schedules by as much as 3.6% in many instances.

In this paper, simulated annealing (SA) is applied to the deterministic dynamic lot-sizing problem with batch ordering and backorders. Batch ordering requires orders that are integer multiples of a fixed quantity that is larger than 1. The performance of the developed SA heuristic is compared to that of a genetic algorithm (GA) and a modified silver-meal (MSM) heuristic developed in the literature, based on the frequency of obtaining the optimum solution and the percentage average deviation from the optimum solution. In addition, the effects of three factors on the performance of the SA, GA, and the MSM are investigated in a 23 factorial experiment. The investigated factors are the demand pattern, the batch size, and the length of the planning horizon. Results indicate that the SA heuristic has the best performance, followed by GA, in terms of the frequency of obtaining the optimum solution and the average deviation from the optimum solution. SA is also the most robust of the investigated heuristics as its performance is only affected by the length of the planning horizon.

In this paper, genetic algorithms are applied to the deterministic time-varying lot sizing problem with batch ordering and backorders. Batch ordering requires orders that are integer multiples of a fixed quantity that is larger than one. The developed genetic algorithm (GA) utilizes a new '012' coding scheme that is designed specifically for the batch ordering policy. The performance of the developed GA is compared to that of a modified Silver-Meal (MSM) heuristic based on the frequency of obtaining the optimum solution and the average percentage deviation from the optimum solution. In addition, the effect of five factors on the performance of the GA and the MSM is investigated in a fractional factorial experiment. Results indicate that the GA outperforms the MSM in both responses, with a more robust performance. Significant factors and interactions are identified and the best conditions for applying each approach are pointed out.

In this paper, genetic algorithms and simulated annealing are applied to scheduling in agile manufacturing. The system addressed consists of a single flexible machine followed by multiple identical assembly stations, and the scheduling objective is to minimize the makespan. Both genetic algorithms and simulated annealing are investigated based on random starting solutions and based on starting solutions obtained from existing heuristics in the literature. Overall, four new algorithms are developed and their performance is compared to the existing heuristics. A 23 factorial experiment, replicated twice, is used to compare the performance of the various approaches, and identify the significant factors that affect the frequency of resulting in the best solution and the average percentage deviation from a lower bound. The results show that both genetic algorithms and simulated annealing outperform the existing heuristics in many instances. In addition, simulated annealing outperforms genetic algorithms with a...

Artificial neural network models have been used successfully to solve demand forecasting and production scheduling problems; the two steps that typically precede and succeed Material Requirements Planning (MRP). In this paper, a neural network model is applied to the MRP problem of lot-sizing. The model’s performance is evaluated under different scenarios and is compared to common heuristics that address the same problem. Results show that the developed artificial neural network model is capable of solving the lot-sizing problem with notable consistency and reasonable accuracy.

This paper presents a methodology for maintaining the operational validity of simulation models of observable systems in order to support operational decisions. In this methodology, real-time system data are continuously compared against simultaneous prediction intervals on selected responses constructed using the simulation model. The methodology is illustrated through using a case example of a simulation model of a flexible manufacturing system. Different invalidating discrepancies between the model and the system are investigated. Results indicate that using nontraditional responses may lead to a faster detection of invalidating changes, the speed of detection is a function of the scope of the change, and the model may evolve with the system and continue to be used to guard against random changes.

The goal of business process reengineering (BPR) is to achieve dramatic improvements in business measures of performance by radically changing the process design. This paper provides a structural definition of the redesign search space using the number of employee types and the number of activities. A systematic approach that combines some linear programming tools with unique BPR principles is then proposed to support the technical redesign of the current process. The approach is illustrated using a public firm process.

Process Reengineering (PR) is one of the most popular work system improvement concepts in industry today. In PR, the basic designs of key organizational processes are reconsidered in light of the tremendous advancements in technology; especially with regard to computer power, software capabilities, and communication. In this paper, PR is presented as a constrained design search problem, where a solution is located using one of two proposed algorithms. The first is an aggregation algorithm which commences the search from the current design of the process, whereas the second is a decomposition algorithm which commences the search from a hypothetical design composed of non-distinct activities. Each algorithm supports a different approach for reconsidering the current process. Aggregation supports orderly elimination of activities from the current process, whereas decomposition supports a novel construction of the reengineered process. Alternative designs are compared using a specially constructed utility function tailored to the objectives of PR. An architecture for a PR decision support system (PRDSS) is presented. PRDSS supports the entire process of generating, evaluating, and selecting the best design using either of the two proposed algorithms. A case example of process reengineering in a public service organization is also included to explain the various concepts presented in the paper. © 1997 Elsevier Science Ltd

In this paper, process reengineering (PR) is presented as a constrained design search problem, where a solution is located using one of two alternative search algorithms. The algorithms are used to develop a framework for a PR decision support system (PRDSS). We believe that the result of our work lays the ground for a more orderly and structured definition and development of the PR concept.

In this paper, a mathematical model to determine the optimal lot size in group technology systems is developed, based on a continuous consumption pattern. The optimal lot size formula that minimizes total cost is then determined. This formula is found to be similar to the standard economic production quantity formula, except for the cost elements associated with items' production and the opportunity cost associated with total machines set-up time. However, it is shown that in practical situations, where machine set-up consumes a limited time, the difference between the two formulas is insignificant. A variant formula for the case where storage cost is based on a fixed charge in a random access storage is also derived. Sensitivity analysis with respect to annual demand rate, annual carrying cost rate, cost of raw material handling, total machine set-up time and cost per unit time is presented. Finally, an example is worked out to illustrate the analytical results.

Automated Optical Inspection (AOI) systems are rapidly replacing slow and tedious manual inspections of Printed Circuit Boards (PCBs). In an AOI system, a minicamera traverses the PCB in a pre-defined travel path, snapping shots of all the PCB components or nodes, at pre-defined locations. The images are then processed and information about the different nodes is extracted and compared against ideal standards stored in the AOI system. This way, a flawed board is detected. Minimizing both the number of images required to scan all the PCB nodes, and the path through which the camera must travel to achieve this, will minimize the image acquisition time and the traveling time, and thus the overall time of inspection. This consequently both reduces costs and increases production rate. This work breaks down this problem into two sub-problems: The first is a clustering problem; the second a travelling salesman sequencing problem. In the clustering problem, it is required to divide all the nodes of a PCB into the minimum number of clusters. The cluster size is constrained by the given dimensions of the camera’s scope or Field of Vision (FOV). These dimensions determine the dimension of the inspection windows. It is thus required to find the minimum number of inspection windows that will scan all the nodes of a PCB, and their locations. Genetic algorithms are applied in a two-step approach with special operators suited for the problem. A continuous Genetic Algorithm (GA) is applied to find the optimum inspection window locations that cover one node and as many other nodes as possible. A discrete GA is then applied to eliminate redundant inspection windows leaving the minimum number of windows that cover all nodes throughout the PCB. In the second sub-problem, an Ant Colony Optimization (ACO) method is used to find the optimum path between the selected inspection windows. The method proposed in this paper is compared against relevant published work, and it is shown to yield better results.Copyright © 2010 by ASME

The research presented in this paper reports on a shop-floor simulation interface (SFSI) which utilizes an integrated bar code-based data acquisition system to improve the real-time performance of an existing simulation-based decision support system and expand its capabilities to shop-floor monitoring and prediction. SFSI utilizes the data acquisition system to detect any shop-floor problem, and to update and initialize simulation models to the current shop-floor status. SFSI can process the simulation model, determine its initial conditions' requirements, retrieve the required information from the data acquisition system, and update the simulation model. This allows the simulation to start with exact current system conditions eliminating the need for time consuming warm-up periods, thus improving the system's accuracy and real-time performance.