Multi-objective Sequencing Problem Research Articles

Approximate Local Search Methods for Multi-objective Sequencing Problem

In the setting of a single machine, this study suggests approximation local search strategies to identify approximate solutions to the multi-objective sequencing problem, where the problem is the total of the four objectives: total completion time ∑Cj j= 1,….,n, total lateness ∑Lj, maximum lateness Lmax and maximum earliness Emax. Descent Method (DM), Simulated Annealing (SA), and genetic algorithm (GA) are three approximate local search techniques that are computer-implemented Matlab programs. On the basis of the outcomes of computing tests, conclusions are modeled on the effectiveness of the local search techniques.

A genetic regulatory network based method for multi-objective sequencing problem in mixed-model assembly lines.

This research proposes a genetic regulatory network based sequencing method that minimizes multiple objectives including utility work costs, production rate variation costs and setup costs in mixed-model assembly lines. After constructing mathematical model of this multi-objective sequencing problem, the proposed method generates a set of genes to represent the decision variables and develops a gene regulation equation to describe decision variable interactions composed of production constraints and some validated sequencing rules. Moreover, a gene expression procedure that determines each gene's expression state based on the gene regulation equation is designed. This enables the generation of a series of problem solutions by indicating decision variable values with related gene expression states, and realizes the minimization of weighted sum of multiple objectives by applying a regulatory parameter optimization mechanism in regulation equations. The proposed genetic regulatory network based sequencing method is validated through a series of comparative experiments, and the results demonstrate its effectiveness over other methods in terms of solution quality, especially for industrial instances collected from a diesel engine assembly line.

Optimizing multi-objective sequencing problem in mixed-model assembly line on just-in-time: particle swarm optimization algorithm

This paper deals with the formulation of a sequencing problem with the dual goals of varying the parts utilization at different workstations of the assembly line and varying the workload associated with each workstation; these two objectives are typically inversely correlated with each other, and therefore the simultaneous optimization of both is challenging. Owing to the NP-hardness of the problem, this paper introduces a Discrete Particle Swarm Optimization (DPSO) algorithm, a Memetic Algorithm (MA), a Weighted sum Multi-Objective Genetic Algorithm (MOGAW), and a Non-dominated Sorting Genetic Algorithm (NSGA-II) to solve a just-in-time sequencing problem where these objectives are to be optimized simultaneously. The performance of these algorithms is compared against each other on small, medium, and large problems. According to statistical analysis, experimental results show that discrete particle swarm optimization outperforms the other algorithms in respect of a comparison metric.

A Hybrid GA-SA Algorithm for Multi-Objective Sequencing Problem in High-Product Mix Shop-Floor

As globalization has increased in the past few years, many companies attempts to made appropriate strategic decision to meet with this challenge. The problem under study mainly focuses on minimizing overall make-span but additional objectives such as balancing the assembly line and minimizing the variation of completion time are also considered. Due to the complexity of problem solving procedure by mathematical techniques, this paper presents a new approach of hybrid GA-SA implementation in order to meet the problem objectives. A proposed hybrid GA-SA is executed to overcome the problem complexity and meet the problem objectives. In order to check the efficiency of hybrid search techniques, a comparison is done between the results obtained by hybrid GA-SA and simple GA and the results comparison validates the effectiveness of presented hybrid search techniques.

A new particle swarm algorithm for a multi-objective mixed-model assembly line sequencing problem

The sequencing of products for mixed-model assembly line in Just-in-Time manufacturing systems is sometimes based on multiple criteria. In this paper, three major goals are to be simultaneously minimized: total utility work, total production rate variation, and total setup cost. A multi-objective sequencing problem and its mathematical formulation are described. Due to the NP-hardness of the problem, a new multi-objective particle swarm (MOPS) is designed to search locally Pareto-optimal frontier for the problem. To validate the performance of the proposed algorithm, various test problems are solved and the reliability of the proposed algorithm, based on some comparison metrics, is compared with three distinguished multi-objective genetic algorithms (MOGAs), i.e. PS-NC GA, NSGA-II, and SPEA-II. Comparison shows that MOPS provides superior results to MOGAs.

A multi-objective scatter search for a mixed-model assembly line sequencing problem

A mixed-model assembly line (MMAL) is a type of production line where a variety of product models similar to product characteristics are assembled. There is a set of criteria on which to judge sequences of product models in terms of the effective utilization of this line. In this paper, we consider three objectives, simultaneously: minimizing total utility work, total production rate variation, and total setup cost. A multi-objective sequencing problem and its mathematical formulation are described. Since this type of problem is NP-hard, a new multi-objective scatter search (MOSS) is designed for searching locally Pareto-optimal frontier for the problem. To validate the performance of the proposed algorithm, in terms of solution quality and diversity level, various test problems are made and the reliability of the proposed algorithm, based on some comparison metrics, is compared with three prominent multi-objective genetic algorithms, i.e. PS-NC GA, NSGA-II, and SPEA-II. The computational results show that the proposed MOSS outperforms the existing genetic algorithms, especially for the large-sized problems.