TY - JOUR

T1 - Genetic algorithms

T2 - A tool for modelling, simulation, and optimization of complex systems

AU - Szczerbicka, Helena

AU - Becker, Matthias

AU - Syrjakow, Michael

PY - 1998/10/29

Y1 - 1998/10/29

N2 - Until very recently genetic algorithms GAs were considered to be the proprietary field of general systems theoreticians and important for esoteric or extremely complex optimization studies. This paper endeavors to show that GA are of great utility in cases where complex systems have to be designed and, therefore, rational choices have to be made. The GA approach is based loosely on the theory of natural evolution, genetic diversity, and searching for beneficial adaptations to a complicated and changing environment. GAs can be viewed as a modelling tool and as a technique for simulation of complex systems represented by communities of interacting units. The representation of units can express characteristics, capabilities, or relatively simple strategies. These units compete and are modified by external operators, so that the overall system adapts to its environment. That environment defines the criterion by which the success in adapting can be measured. Genetic algorithms have been successfully applied to many optimization problems including mathematical function optimization, very large scale integration VLSI chip layout, molecular docking, parameter fitting, scheduling, manufacturing, clustering, machine learning, etc. and are still finding increasing acceptance. Modelling and optimization of a Kanban system from the field of flexible manufacturing systems is discussed in the last section.

AB - Until very recently genetic algorithms GAs were considered to be the proprietary field of general systems theoreticians and important for esoteric or extremely complex optimization studies. This paper endeavors to show that GA are of great utility in cases where complex systems have to be designed and, therefore, rational choices have to be made. The GA approach is based loosely on the theory of natural evolution, genetic diversity, and searching for beneficial adaptations to a complicated and changing environment. GAs can be viewed as a modelling tool and as a technique for simulation of complex systems represented by communities of interacting units. The representation of units can express characteristics, capabilities, or relatively simple strategies. These units compete and are modified by external operators, so that the overall system adapts to its environment. That environment defines the criterion by which the success in adapting can be measured. Genetic algorithms have been successfully applied to many optimization problems including mathematical function optimization, very large scale integration VLSI chip layout, molecular docking, parameter fitting, scheduling, manufacturing, clustering, machine learning, etc. and are still finding increasing acceptance. Modelling and optimization of a Kanban system from the field of flexible manufacturing systems is discussed in the last section.

UR - http://www.scopus.com/inward/record.url?scp=0032186746&partnerID=8YFLogxK

U2 - 10.1080/019697298125461

DO - 10.1080/019697298125461

M3 - Article

AN - SCOPUS:0032186746

VL - 29

SP - 639

EP - 659

JO - Cybernetics and Systems

JF - Cybernetics and Systems

SN - 0196-9722

IS - 7

ER -