Multi-objective Optimization of Cloud Manufacturing Service Composition with Cloud-Entropy Enhanced Genetic Algorithm

This paper presents an effective use of hybrid metaheuristics algorithm for solving Job Shop Scheduling Problem (JSSP). Integration of three metaheuristics algorithms: Shuffled Frog Leaping Algorithm (SFLA), Intelligent Water Drops algorithm (IWD) and Path Relinking (PR) algorithm were put together to solve JSSP. First, simulation model was developed and tested on the test data of Traveller Salesman Problem (TSP). Second, the model was tested on real world production line to solve the problem of Minimum Needed Workers (MNW) at the production line. The model enables individual test of three mentioned algorithms and calculation of new proposed Random MultiNeighbourhood based Shuffled Frog Leaping Algorithm with Path Relinking (RMN-SFLA-PR). Experiments were tested on two software environments MATLAB and Simio, which gives us reliable, robust and tangible results. Results show that the new proposed RMN-SFLA-PR algorithm converged to optimum almost ten times faster than individual algorithms. The most important thing is the successful rate of all independent runs of the proposed RMN-SFLA-PR is 100 % in low-dimensional cases of the 4 benchmarks (dj38) and in JSSP to solve MNW for the real world production line.

Section: Literature Reviewmentioning

confidence: 99%

An Effective Use of Hybrid Metaheuristics Algorithm for Job Shop Scheduling Problem

Zhang¹,

Liu²,

Moraca³

et al. 2016

“…The genetic algorithms are population-based search heuristics which simulate the natural evolution of living beings [28] and can be used for solving different problems (e.g., [29][30][31][32]). In general, any other search strategy can be used, such as particle swarm optimization, simulated annealing, bat algorithm, etc.…”

Section: Search Strategymentioning

confidence: 99%

A Universal CAD System for Cutting Stock Problem

Kovačić¹,

Brezočnik²

2017

In this research the universal system using CAD geometry for solving of cutting stock problem is proposed. The system consists of the three main modules: geometry definition module, objective definition module, and the search strategy module. In the first module the reference points and orientation of parts and stock are defined. In the second module some constraints such as the quantity of the desired type of parts to be placed on the stock can be taken into account. The third module contains the search strategy. For the purpose of this research, the genetic algorithm was used as the search method. For illustration of the generality of the proposed approach, three practical examples for solving orthogonally and irregularly-shaped cutting stock problems are presented. The parts which are to be cut from stock-material with minimal loss and also the stock-material itself are presented as CAD geometries without any kind of geometrical constraints (i.e., concavity, convexity, sections, and holes). In the presented approach the AutoCAD environment was used. The AutoLISP in-house developed system for cutting stock problem was integrated into the AutoCAD. The developed system is without any kind of geometrical constraints and thus highly applicable in the practice. The presented approach can be developed also in other CAD systems where the application programming interfaces (API) are available.

“…Not only is it difficult because of its dynamic nature, but it is also considered of little relevance in classical assembly line manufacturing environments, where more interest is shown in machine capacity than in the previous and subsequent movements of both future vehicles and their components. However, the variety of information interaction and material transportation in current Manufacturing Systems (MS) [4], which involves up to five car models on the same line in automobile assembly lines, hinder the analysis of this relevant flow. For that purpose, the development of a virtual layout using simulation modelling is worthwhile during the decision-making process of MS reconfigurations [5].…”

Section: Introductionmentioning

confidence: 99%

Using 4-Layer Architecture to Simulate Product and Information Flows in Manufacturing Systems

Saez-Mas¹,

García-Sabater²,

Morant-Llorca³

2018

This work illustrates the application of novel simulation architecture with two case studies where the proposed architecture, the so-called 4-layer, allowed us to address the complexity of the analysed systems. The fundamental objective of this work is to show the structure of layers, how layers interact with one another and with the user, and what benefits this separation proposes. The first case study deals with moving car bodies from the paint plant to the assembly line through a sequencing system that involves distributed decision-making processes in an ASRS. The second case study focuses on analysing a layout of a section used to assemble the engine and transmission set, where the quality of the material flow is evaluated. The work highlights some of the advantages of modelling with 4-layer architecture, and explains the key processes that connect different elements.