Elsevier

Applied Soft Computing

Volume 5, Issue 3, March 2005, Pages 281-299
Applied Soft Computing

Evolutionary computing in manufacturing industry: an overview of recent applications

https://doi.org/10.1016/j.asoc.2004.08.003Get rights and content

Abstract

Traditional methods often employed to solve complex real world problems tend to inhibit elaborate exploration of the search space. They can be expensive and often results in sub-optimal solutions. Evolutionary computation (EC) is generating considerable interest for solving real world engineering problems. They are proving robust in delivering global optimal solutions and helping to resolve limitations encountered in traditional methods. EC harnesses the power of natural selection to turn computers into optimisation tools. The core methodologies of EC are genetic algorithms (GA), evolutionary programming (EP), evolution strategies (ES) and genetic programming (GP). This paper attempts to bridge the gap between theory and practice by exploring characteristics of real world problems and by surveying recent EC applications for solving real world problems in the manufacturing industry. The survey outlines the current status and trends of EC applications in manufacturing industry. For each application domain, the paper describes the general domain problem, common issues, current trends, and the improvements generated by adopting the GA strategy. The paper concludes with an outline of inhibitors to industrial applications of optimisation algorithms.

Introduction

Real world engineering can be characterised as having chaotic disturbances, randomness and complex non-linear dynamics [1]. Most industrial processes are usually large scale, highly dimensional, non-linear, highly uncertain with complex highly skilled operators to control the process plants. Conventional methods such as trial and error are often used to solve complex optimisation problems. This approach relies on the use of the analyst's qualitative (QL) knowledge to explore the design space [2], [3]. Expensive advanced computational analyses (such as finite element) are also used to understand the behaviour of complex engineering problems. These are often invoked repeatedly during the search process making the optimisation and concept exploration time consuming. This traditional search method often results in sub-optimal solutions due to inherent limitations in incomplete knowledge representation and the fact that elaborate exploration of the design space is inhibited. There is also a tendency to accept local optimal solutions considered to be sufficiently good for the chosen objective due to the following reasons: subjective judgment, similarity to historical result or simply constraint on time to deliver workable solutions. Multiple global optimal solutions are desirable for these classes of problems to give alternative solutions in the presence of increasing dynamic and ill-defined problem space. EC techniques are receiving an increasing interest for solving real world engineering problems. These techniques are proving robust in delivering global optimal quality solutions and are helping to resolve some of the complexity issues encountered in real world problems.

EC is a method of harnessing the power of natural selection to turn computers into optimisation tools. EC is one of the main constituents of soft computing (SC), where SC is a collection of methodologies including neural computing (NC) and fuzzy computation (FC), EC and their various combinations [4]. EC techniques have been successfully applied in many areas including: engineering design optimisation, manufacturing system, process control, medical-diagnosis, and simulation and communication systems.

This paper attempts to bridge the gap between theory and practice by exploring the features of classical optimisation algorithms, EC techniques and characteristics of real world problems. The paper surveys recent innovative EC applications in the manufacturing industry in the following areas: metal forming, chemical industry, paper industry, scheduling and process planning, CAD/CAM environment and manufacturing related industries. For each application area, the paper describes the general domain problem, common issues, current trends, and the improvements generated by adopting the EC strategy. All the application problems that are discussed in this paper are used/piloted in the specific industry. The paper concludes with an outline of inhibitors to industrial applications of optimisation algorithms.

Papers from related IEEE Transactions using IEEE Xplore and other related journals over the last 5 years are reviewed here to show reported industrial innovations using EC. The paper is organised as follows: Section 2 describes the algorithmic approaches to optimisation. EC and its components are described in Section 3. The characteristics of real world problems are explored in Section 4. Industrial applications of EC techniques are discussed in Section 5, Section 6 outlines the inhibitors to industrial applications of EC-based techniques, and the concluding remark is given in Section 7.

Section snippets

Algorithmic approaches to optimisation

This section explores the features of classical optimisation algorithms in order to appreciate the difficulties recent EC-based approaches have to overcome to be successful. Literature suggests a number of optimisation techniques for solving optimisation problems. These techniques can be classified into two broad categories: classical and evolutionary.

Most classical algorithms use a point-by-point deterministic procedure for approaching the optimum solution. Such algorithms start from a random

Evolutionary computation

Evolutionary computing techniques are robust and offer exceptional adaptive capabilities to handle non-linear, highly dimensional and complex engineering problems. They do not require explicit knowledge of the problem structure or differentiability, and have ability to provide multiple near-optimal solutions to even ill-defined problems from the expert. The origins of EC can be traced back to 1950s [14], [15], since then several evolutionary algorithms have been proposed, some of which

Characteristics of real world problems

Most real world manufacturing problems are usually large scale, highly dimensional, non-linear and highly uncertain involving interaction with engineers and highly skilled operators that control the process plants. These problems are also characterised by chaotic disturbances, randomness and complex non-linear dynamics [1]. An understanding of the features of real world problems supports algorithmic development to include wider applications of EC-based algorithms in industry. Therefore some of

Evolutionary computation in industry

EC techniques are gaining increased interest in industry. Literature reveals a number of real-life applications of optimisation algorithms, especially EC. Fig. 3 presents a pie chart showing the proportion for different EC techniques in the sample of publications reviewed in this paper. The sampling is not exhaustive and is just intended to give a snap shot of the research efforts for various EC techniques.

The chart in Fig. 3 is fairly intuitive, with each sector representing the percentage

Inhibitors to industrial applications of EC-based optimisation algorithms

Roy et al. [85] carried out a survey of existing literature in the area of real-life optimisation in order to analyse the state-of-the-art evolutionary-based optimisation algorithms and their real-life applications. This was complimented by an industrial survey in which the designers were interviewed using a semi-formal questionnaire as a support tool. This survey enabled the identification of the features of real-life optimisation problems, current status of optimisation in industry and

Concluding remarks

The successful applications of evolutionary computing (EC) suggest that EC will have increasing impact in future. Evolutionary computing is already having an important role on many industrial operations. It provides alternative approaches to traditional analytical problem solving methods (the fundamental limitations of these classical approaches are discussed in Section 2) and it overcomes their shortcomings by its parallel adaptive nature. The main components (GA, GP, EP and ES) of EC were

References (100)

  • M. Yokoyama et al.

    Optimization of the stochastic dynamic production cycling problem by a genetic algorithm

    Comput. Oper. Res.

    (2003)
  • M. Gen et al.

    Network design techniques using genetic algorithms

    Adv. Eng. Software

    (2001)
  • S.D. Harris et al.

    The optimisation of reaction rate parameters for chemical kinetic modelling of combustion using genetic algorithms

    Comput. Meth. Appl. Mech. Eng.

    (2000)
  • J.K. Rajesh et al.

    Multi-objective optimization of industrial hydrogen plants

    Chem. Eng. Sci.

    (2001)
  • M. Grujicic et al.

    Optimization of the chemical vapor deposition process for carbon nanotubes fabrication

    Appl. Surf. Sci.

    (2002)
  • L. Balland et al.

    Kinetic parameter estimation of solvent-free reactions: application to esterification of acetic anhydride by methanol

    Chem. Eng. Process.

    (2002)
  • R. Luus

    Application of dynamic programming to differential-algebraic process systems

    Comput. Chem. Eng.

    (1993)
  • G. Liden et al.

    Fairing of surfaces with optimization techniques using FANGA curves as quality criterion

    Comput. Aided Des.

    (1993)
  • H. Einar et al.

    FORMELA: a general design and production data system for sculptured products

    Computer Aided Des.

    (1973)
  • M. Szilvasi-Nagy

    Detecting regions with undesirable curvature

    Comput. Aided Geom. Des.

    (1997)
  • H.S. Cho et al.

    Determining optimal parameters for stereolithography processes via genetic algorithm

    J. Manuf. Syst.

    (2000)
  • K. Deb et al.

    Design of truss-structures for minimum weight using genetic algorithms

    Finite Elements Anal. Des.

    (2001)
  • T. Chen et al.

    Determination of optimum design spaces for topology optimization

    Finite Elements Anal. Des.

    (2000)
  • D.B. Fogel

    What is Evolutionary computation?

    IEEE Spectrum

    (2000)
  • R. Roy, Adaptive search and the preliminary design of gas turbine blade cooling system, Ph.D. Thesis, University of...
  • V. Oduguwa, R. Roy, Qualitative and quantitative knowledge in engineering system design, Presented at the 17th National...
  • A. Tettamanzi et al.

    Soft Computing: Integrating Evolutionary, Neural and Fuzzy System

    (2001)
  • K. Deb

    Optimization for Engineering Design: Algorithms and Examples

    (1995)
  • K. Deb

    Multi-objective Optimization using Evolutionary Algorithms

    (2001)
  • W. Spendley et al.

    Sequential application of simplex designs in optimization and evolutionary operation

    Technometrics

    (1962)
  • R. Hooke et al.

    Direct search solution of numerical and statistical problems

    J. ACM

    (1961)
  • M.J.D. Powell

    An efficient method for finding the minimum of a function of several variables without calculating derivatives

    Comput. J.

    (1964)
  • A.L. Cauchy

    Methode generale pour la resolution des systemes d’equations simultanees

    Comptes rendus de l’academie des sciences

    (1847)
  • D. Marquardt

    An algorithm for least squares estimation of non-linear parameters

    SIAM J. Appl. Math.

    (1963)
  • R. Fletcher et al.

    Function minimization by conjugate gradients

    Comput. J.

    (1964)
  • A. Tiwari, Evolutionary computing techniques for handling variable interaction in engineering design optimisation,...
  • G.E.P. Box

    Evolutionary operation: a method of increasing industrial productivity

    Appl. Stat.

    (1957)
  • A.S. Fraser

    Simulation of genetic systems by automatic digital computers. I. Introduction

    Aust. J. Biol. Sci.

    (1957)
  • J.H. Holland

    Outline of a logical theory of adaptive systems

    J. ACM

    (1962)
  • J.H. Holland

    Adaptation in Natural and Artificial Systems

    (1975)
  • J.R. Koza

    Genetic Programming

    (1992)
  • L.J. Fogel et al.

    Artificial Intelligence through Simulated Evolution

    (1966)
  • I. Rechenberg, Cybernetic solution path of an experimental problem, Royal Aircraft Establishment, Library Translation...
  • H.-P. Schwefel

    Numerical Optimisation of Computer Models

    (1981)
  • D.B. Fogel

    Evolutionary Computation

    (1995)
  • K.E. Kinnear

    Advances in Genetic Programming.

    (1994)
  • T. Back, H.-P. Schwefel, Evolutionary computation: an overview, Presented at the Proceedings of the Third IEEE...
  • D.H. Wolpert

    No Free Lunch Theorems for Optimization

    (1996)
  • H.W. Ray et al.

    Process Optimization with Applications in Metallurgy and Chemical Engineering

    (1973)
  • R. Bellman

    Dynamic Programming

    (1957)
  • Cited by (129)

    View all citing articles on Scopus
    View full text