Multiple Criteria Transportation Application With Fuzzy Cost Parameters Using Genetic Algorithm

Paryati, Paryati (2021) Multiple Criteria Transportation Application With Fuzzy Cost Parameters Using Genetic Algorithm. In: 2021 Asia Pacific Industrial Engineering & Management Systems 5th Webinar, 3-4 Dec 2021, Taiwan.

Text PARYATI Proceedings APIEMS 2021.pdf Download (435kB)

Abstract

The design and implementation of a software used as a tool aid to create a multiple criteria transportation model which is equipped with a fuzzy cost parameter by using genetic algorithm, has been made. This software is called TraFAG.
Waterfall methodology, which comprised of analyzing, designing, implementing and testing processes, was used in the software engineering. The algorithm used in these processes is a genetic algorithm. It is based on the genetic processed of living creatures, that is, generation processes in a natural population which ultimately follow selection principles, or where it is only the strong that will survive. In the transportation system, the impacts of transportation causes uncertainty on some or all coefficients of the objective functions, such as transportation costs or delivery time becomes unclear. One way to deal with uncertainty in making such decision is by using fuzzy principles. The parameters of the fuzzy costs on TraFAG uses Triangular Fuzzy Number (TFN). In multiple criteria optimizations, the decision of optimum value uses Pareto solution. Pareto Solution is determined on the basis of ordered values of fuzzy destination. The comparison and order of the fuzzy values uses integral values. The TraFAG software is applied in the programming language environment of Borland Delphi Version 7.0 one that is developed from Pascal for Window based programming environment.
The solution of multiple criteria transportation problem can be solved by heuristic approach using genetic algorithm. The analysis of programme value shows that the process on evaluation case will straight proportional with the result of the multiplication of the source of depot total and the destination depot with the corelation coefisien 0.93. The analysis also shows that the amount of population is straight proportional/linear to each of the case evaluation towards the process time with correlation coefisien 0.96. Parameter α shows the optimism degree will influence the result of the integral value linearly. The higher of the α value, so the cost of transportation is the bigger. It is better to choose the α with has value 0.5, which has moderate value in order that it will be in safe condition. The α which produces minimum cost for the evaluation case 2 up to 6 is 0.1. The bigger of the population amount has inclination smaller of its fitness function. The bigger amount of the generation so the transportation cost will be smaller. The value which is gotten is in relative average stable in the generation over 500. The crossover probability influences to the fitness function. In case 15 and 17 the crossover probability causes the value of the fitness function minimum 0.1. Transfer transportation has more influence to the fitness function. In case 2 causes the transfer transportation stable with the fitness value 47.65.

Item Type:	Conference or Workshop Item (Paper)
Additional Information:	Publish Conference International
Uncontrolled Keywords:	Genetic Algorithm, Fuzzy Logic, Transportation Problem, Waterfall, Multiple criteria
Subjects:	T Technology > T Technology (General)
Divisions:	Faculty of Engineering, Science and Mathematics > School of Engineering Sciences
Depositing User:	ST.,M.Kom PARYATI PARYATI
Date Deposited:	07 Apr 2023 09:18
Last Modified:	07 Apr 2023 09:27
URI:	http://eprints.upnyk.ac.id/id/eprint/33603

Actions (login required)

View Item