Romeo Marian

A genetic algorithm for the optimisation of assembly sequences

This paper describes a Genetic Algorithm (GA) designed to optimise the Assembly Sequence Planning Problem (ASPP), an extremely diverse, large scale and highly constrained combinatorial problem. The modelling of the ASPP problem, Which has to be able to encode any industrial-size product with realistic constraints, and the GA have been designed to accommodate any type of assembly plan and component. A number of specific modelling issues necessary for understanding the manner in which the algorithm works and how it relates to real-life problems, are succinctly presented, as they have to be taken into account/adapted/solved prior to Solving and Optimising (S/O) the problem. The GA has a classical structure but modified genetic operators, to avoid the combinatorial explosion. It works only with feasible assembly sequences and has the ability to search the entire solution space of full-scale, unabridged problems of industrial size. A case study illustrates the application of the proposed GA for a 25-components product.

Assembly sequence planning and optimisation using genetic algorithms: Part I. Automatic generation of feasible assembly sequences

Abstract This paper attempts to formalise, solve and optimise (S/O) the Assembly Sequence Planning Problem (ASPP), a large scale, highly constrained combinatorial problem. Due to the complexity of the subject and the number of related matters to be considered/adapted/solved prior to S/O the ASPP, the paper is split in two, self-contained, parts: Part I—Automatic Generation of Feasible Assembly Sequences and Part II—optimisation of assembly sequences using Genetic Algorithms. The first part deals with formalising the ASPP—modelling and representation issues—and generating feasible assembly sequences (solving the ASPP). The second part is concerned with the optimisation of the ASPP and will present in detail the Genetic Algorithm designed to optimise it, the genetic operators that compose the algorithm and the definition of the fitness function (optimisation function). The ASPP is considered here as a full-scale, unabridged problem.

Genetic algorithm optimisation of an integrated aggregate production–distribution plan in supply chains

A production plan concerns the allocation of resources of the company to meet the demand forecast over a certain planning horizon and a distribution plan involves the management of warehouse storage assignments, transport routings and inventory management issues. A production–distribution plan integrates the decisions in production, transport and warehousing as well as inventory management. The overall performance of a supply-chain is influenced significantly by the decisions taken in its production–distribution plan and hence one key issue in the performance evaluation of a supply chain is the modelling and optimisation of the production–distribution plan considering its actual complexity. Based on the integration of Aggregate Production Plan and Distribution Plan, this article develops a mixed integer non-linear formulation for a two-echelon supply network (i.e. a production-distribution network) considering the real-world variables and constraints. Genetic Algorithm (GA), known as a robust technique for solving complex problems, is employed for the optimisation of the developed mathematical model due to its ability to effectively deal with a large number of parameters. To demonstrate the applicability of the methodology, a real-life case study will be finally studied incorporating the production of different types of products in several manufacturing plants and the distribution of finished products from plants to a number of end-users via multiple direct/indirect transport routes.

Optimisation of partner selection and collaborative transportation scheduling in Virtual Enterprises using GA

Partner selection and transportation scheduling are critical to the success of a Virtual Enterprise. Collaborative transportation is a promising strategy that can help many enterprises survive and thrive in todays highly competitive market. To help decision makers establish and operate Virtual Enterprises more effectively, an innovative decision support system is proposed in this paper. First, new model for integration of partner selection and collaborative transportation scheduling in Virtual Enterprises is developed. This integrated optimisation problem is very dynamic in nature and it is required to optimise a number of interlinked sub-problems at the same time. Then, a novel Genetic Algorithm with a unique dynamic chromosome representation and genetic operations is developed to find an optimal solution to the integrated problem. The effectiveness of the proposed approach is demonstrated in a representative case study.

Simulation modelling and analysis of scheduling in robotic flexible assembly cells using Taguchi method

Due to increasing competition in the developing global economy, today’s companies are facing greater challenges than ever to employ flexible manufacturing systems (FMS) capable of dealing with unexpected events and meeting customers’ requirements. One such system is robotic flexible assembly cells (RFACs). There has been relatively little work on the scheduling of RFACs, even though overall scheduling problems of FMS have attracted significant attention. This paper presents Taguchi optimisation method in conjunction with simulation modelling in a new application for dynamic scheduling problems in RFACs, in order to minimise total tardiness and number of tardy jobs (NT). This is the first study to address these particular problems. In this study, Taguchi method has been used to reduce the minimum number of experiments required for scheduling RFACs. These experiments are based on an L9 orthogonal array with each trial implemented under different levels of scheduling factors. Four factors are considered simultaneously: sequencing rule, dispatching rule, cell utilisation and due date tightness. The experimental results are analysed using an analysis of mean to find the best combination of scheduling factors and an analysis of variance to determine the most significant factors that influence the system’s performance. The resulting analysis shows that this proposed methodology enhances the system’s scheduling policy.

Analysing the hindrances to the reduction of manufacturing lead-time and their associated environmental pollution

Manufacturing Lead-Time (MLT) is the total time required to process a given product through a plant. Long MLT is the major cause of inefficient manufacturing, since it generates large amount of wastes and creates considerable environmental burden. In the past, a large amount of environmental pollution has been generated by manufacturing industries, most of which come from inefficient practices trough generating various types of wastes. The easiest and probably the least expensive way to cut manufacturing wastes and minimise their consequent environmental impacts is to improve the manufacturing procedures of the companies. A novel approach to do this is to identify and reduce/eliminate the processes and activities which cause the inefficient use of resources. This paper presents the MLT reduction process via identifying the value-adding and non-value adding activities. It also analyses four crucial sources of wastes which generate the greatest environmental pollution and suggestions for improvements are provided.

Optimization of a Two-Echelon Supply Network Using Multi-objective Genetic Algorithms

The overall performance of a supply-chain (SC) is influenced significantly by the decisions taken in its production-distribution (P-D) plan. A P-D plan integrates decisions in production, transport and warehousing as well as inventory management. One key issue in the performance evaluation of a Supply Network (SN) is the modeling and optimization of P-D planning problem considering its actual complexity. Based on the integration of Aggregate Production Planning and Distribution Planning, this paper firstly develops a mixed integer formulation for a two-echelon supply network considering the real-world variables and constraints. A multi-objective genetic algorithm (MOGA) is then designed for the optimization of the developed mathematical model. Finally, a real-world case study incorporating multiple products, multiple plants, multiple warehouses, multiple end-users, and multiple time periods will be considered for investigating the performance evaluation of the MOGA method against the traditional approaches of SC planning.

Real Coded Genetic Algorithms for Solving Flexible Job-Shop Scheduling Problem - Part II: Optimization

This paper addresses optimization of the flexible job-shop problem (FJSP) by using real-coded genetic algorithms (RCGA) that use an array of real numbers as chromosome representation. The first part of the papers has detailed the modelling of the problems and showed how the novel chromosome representation can be decoded into solution. This second part discusses the effectiveness of each genetic operator and how to determine proper values of the RCGAs parameters. These parameters are used by the RCGA to solve several test bed problems. The experimental results show that by using only simple genetic operators and random initial population, the proposed RCGA can produce promising results comparable to those achieved by other best-known approaches in the literatures. These results demonstrate the robustness of the RCGA.

Optimisation of distribution networks using Genetic Algorithms. Part 2 ? the Genetic Algorithm and Genetic Operators

This paper presents a methodology developed for the optimisation of distribution networks based on Genetic Algorithms (GA), specifically capacitated location-allocation problems. Due to the complexity and extent of the paper, it was split into two parts. Modelling issues and automatic generation of initial population of chromosomes were treated in the first part. This second part details the rest of the GA. Due to the intricacy of the problem, the GA was designed to work only with feasible chromosomes and modified operators were chosen to handle its highly constrained character. They are presented in detail. An example of applying the algorithm for 25 Production Facilities (PF), 10 Distribution Centres (DCs) and 25 Retailers (R) – including 520 variables, tightly interconnected – is presented, demonstrating the robustness of GA and its capacity to tackle problems of considerable size.

Optimization of part type selection and loading problem with alternative production plans in flexible manufacturing system using hybrid genetic algorithms - part 1: Modelling and representation

Two NP-hard and strongly related problems in flexible manufacturing system (FMS), part type selection problem and loading problem, are addressed in this paper. Various flexibilities including alternative production plans are considered. This effort will further exploit the flexibility of the FMS and improve system productivity. Real coded genetic algorithms (RCGA) which uses an array of real numbers as chromosome representation is proposed to solve these problems. Hybridizing the RCGA with variable neighborhood search (VNS) is performed to obtain better results. A strategy to maintain population diversity and avoid a premature convergence is also implemented. This first part of the paper addresses a modeling of the problems and discusses how the chromosome representation of the RCGA can handle various flexibilities of operations in the FMS. The second part of the paper will discuss the effectiveness of this hybrid approach to solve several test bed problems.