Tamer F. Abdelmaguid

A genetic algorithm approach to the integrated inventory-distribution problem

We introduce a new genetic algorithm (GA) approach for the integrated inventory distribution problem (IIDP). We present the developed genetic representation and use a randomized version of a previously developed construction heuristic to generate the initial random population. We design suitable crossover and mutation operators for the GA improvement phase. The comparison of results shows the significance of the designed GA over the construction heuristic and demonstrates the capability of reaching solutions within 20% of the optimum on sets of randomly generated test problems.

A hybrid GA/heuristic approach to the simultaneous scheduling of machines and automated guided vehicles

In this paper, the problem of simultaneous scheduling of machines and identical automated guided vehicles (AGVs) in flexible manufacturing systems is addressed with the objective of minimizing the makespan. This problem is composed of two interrelated decision problems: the scheduling of machines, and the scheduling of AGVs. Both problems are known to be NP-complete, resulting in a more complicated NP-complete problem when they are considered simultaneously. A new hybrid Genetic-algorithm/heuristic coding scheme is developed for the studied problem. The developed coding scheme is combined with a set of genetic algorithm (GA) operators selected from the literature of the applications of GAs to the scheduling problems. The algorithm is applied to a set of 82 test problems, which was constructed by other researchers, and the comparison of the results indicates the superior performance of the developed coding.

Heuristic approaches for the inventory-routing problem with backlogging

We study an inventory-routing problem in which multiperiod inventory holding, backlogging, and vehicle routing decisions are to be taken for a set of customers who receive units of a single item from a depot with infinite supply. We consider a case in which the demand at each customer is deterministic and relatively small compared to the vehicle capacity, and the customers are located closely such that a consolidated shipping strategy is appropriate. We develop constructive and improvement heuristics to obtain an approximate solution for this NP-hard problem and demonstrate their effectiveness through computational experiments.

Representations in Genetic Algorithm for the Job Shop Scheduling Problem: A Computational Study

Due to the NP-hardness of the job shop scheduling problem (JSP), many heuristic approaches have been proposed; among them is the genetic algorithm (GA). In the literature, there are eight different GA representations for the JSP; each one aims to provide subtle environment through which the GA’s reproduction and mutation operators would succeed in finding near optimal solutions in small computational time. This paper provides a computational study to compare the performance of the GA under six different representations.

A dynamic programming approach for minimizing the number of drawing stages and heat treatments in cylindrical shell multistage deep drawing

Deep drawing is an important sheet metal forming process that appears in many industrial fields. It involves pressing a blank sheet against a hollow cavity that takes the form of the desired product. Due to limitations related to the properties of the blank sheet material, several drawing stages may be needed before the required shape and dimensions of the final product can be obtained. Heat treatment may also be needed during the process in order to restore the formability of the material so that failure is avoided. In this paper, the problem of minimizing the number of drawing stages and heat treatments needed for the multistage deep drawing of cylindrical shells is addressed. This problem is directly related to minimizing manufacturing costs and lead time. It is required to determine the post-drawing shell diameters along with whether heat treatment is to be conducted after each drawing stage such that the aforementioned objectives are achieved and failure is avoided. Conventional computer-aided process planning (CAPP) rules are used to define the search space for a dynamic programming (DP) approach in which both the post-drawing shell diameter and material condition are used to define the states in the problem. By discretizing the range of feasible shell diameters starting from the initial blank diameter down to the final shell diameter, the feasible transitions from state to another is represented by a directed graph, based upon which the DP functional equation is easily defined. The DP generates a set of feasible optimized process plans that are then verified by carrying out finite element analysis in which the deformation severity and the resulting strains and thickness variations are investigated. Two case studies are presented to demonstrate the effectiveness of the developed approach. The results suggest that the proposed approach is a valuable, reliable and quick computer aided process planning approach to this complicated problem.

A neighborhood search function for flexible job shop scheduling with separable sequence-dependent setup times

A neighborhood search function for the flexible job shop scheduling with sequence-dependent setup times is developed.Experimentations are conducted to determine its best working parameters under different problem conditions.A tabu search approach is developed to demonstrate the utilization of the developed neighborhood search function.Simple lower bounds are introduced to assess the performance of the developed tabu search approach. This paper addresses the makespan minimization problem in scheduling flexible job shops whenever there exist separable sequence-dependent setup times. An extension to the neighborhood search functions of Mastrolilli and Gambradella, developed for the flexible job shop scheduling problem (FJSP), is provided. It is shown that under certain conditions such an extension is viable. Accordingly, a randomized neighborhood search function is introduced, and its best search parameters are determined experimentally using modified FJSP benchmark instances. A tabu search approach utilizing the proposed neighborhood search function is then developed, and experimentations are conducted using the modified instances to benchmark it against a lower bound. Experimental results show that on average, the tabu search approach is capable of achieving optimality gaps of below 10% for instances with low average setup time to processing time ratios.

A hybrid PSO-TS approach for proportionate multiprocessor open shop scheduling

In this paper, a hybrid particle swarm optimization (PSO)-tabu search (TS) approach is proposed for solving the proportionate multiprocessor open shop scheduling problem (PMOSP) with the objective of minimizing the makespan. The PSO part of the proposed approach is used for randomly searching the machine selection decisions, while the TS part conducts local improvements for the routing and sequencing subproblems. Experimentations are conducted on 100 benchmark problems which are divided into four equal sets with 2, 4, 8 and 16 processing centers. The analysis shows that the proposed hybrid approach produces competitive results compared to previously developed TS and genetic algorithm approaches, especially for intermediate size problems of 4 and 8 processing centers. The average optimality gap of the proposed approach is found to be below 5.6% from the lower bound for the four sets, and ten new upper bounds are found, among them two are provably optimal.

Halting decisions for gas pipeline construction projects using AHP: a case study

This paper considers a decision making problem encountered by a natural gas pipeline construction company having a set of ongoing projects and facing unpredictable risks that can result in large deviations from planned schedules. This situation forces the company to consider the decision of halting one or more projects to avoid future losses and to allow for possible reallocation of some of their resources to other ongoing projects. This decision making problem involves different factors and criteria that need to be combined in an organized structure that exploits assessments of experts managing such projects. The analytic hierarchy process (AHP) is found to be suitable for guiding decisions in this problem. A case study for a major natural gas pipeline construction company in Egypt is presented, where three ongoing projects are considered. The proposed AHP structure, along with collected pairwise comparison scores and calculated priorities, suggests halting one project. Sensitivity analysis is conducted to investigate the effect of changes in the pairwise comparison scores assigned to the main criteria on the final decision. The results and analysis provide some insights regarding the application of the AHP and the relative importance of the factors affecting decisions.

Using NSGA-II to optimise tool life and production time for turning under Minimum Quantity Lubrication

Metal Working Fluids (MWFs) are known to improve machining performance, yet they have poor ecological and health side effects. Therefore, eliminating or reducing their quantity in machining operations is crucial. The Minimum Quantity Lubrication (MQL) is a new sustainable manufacturing technique that can achieve significant reduction in the MWF used compared to traditional wet flooding, while maintaining high performance. This paper provides an experimental investigation to study the characteristics of the flow of the MWF in a turning process utilising the MQL technique and to analyse the effect of the WMF’s behaviour on cutting force, surface roughness and tool wear. Several experiments are conducted considering different workpiece materials and cutting parameters. Based on the experimental results, the Response Surface Methodology (RSM) is used to provide mathematical models that relate the main cutting parameters, the workpiece material properties and the MWF viscosity and flow rate with cutting force, surface roughness and tool wear. A non-linear, multi-objective optimisation problem is formulated for a case study with the objectives of minimising production time and maximising tool life. It is demonstrated that the second version of the Non-dominated Sorting Genetic Algorithm (NSGA-II) is an efficient technique for generating a set of well-spread Pareto front solutions, which helps in determining the most appropriate values of MQL and cutting parameters.

An Integrated Approach for Optimized Process Planning of Multistage Deep Drawing

This work is concerned with the process design of multistage deep drawing, where an integrated artificial intelligence (AI) approach is presented with a special focus on box-shaped parts. This approach combines three AI tools, namely part shape recognition, expert system for process design governing rules, and search and optimization via dynamic programming. Validation and final selection of optimized process plans are done using finite element analysis with full account of the formability limits of the material used. The main advantage of the proposed integrated approach is its capability of generating valid, optimized process plans in a relatively short time compared to traditional approaches. Two case studies are presented for demonstrating its effectiveness.