Evolucijski algoritem za optimiranje prevoza tovora med dvema lokacijama s skupino vozil

Abstract

Keywords

Data

Language:	Slovenian
Year of publishing:	2008
Typology:	2.11 - Undergraduate Thesis
Organization:	UL FRI - Faculty of Computer and Information Science
Publisher:	[E. Dovgan]
UDC:	004.8:656
COBISS:	22018343
Views:	1141
Downloads:	300
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	[Evolutionary algorithm for optimization of cargo transport between two locations with a group of vehicles]
Secondary abstract:	This thesis deals with the optimization of heterogeneous cargo transport between two locations with a group of vehicles. With the increasing size the problem becomes too complex to be solved with deterministic algorithms. Therefore we designed an evolutionary algorithm which belongs to the family of stochastic algorithms. This kind of algorithms is used for solving problems whose solving time with deterministic algorithms would be unacceptable. A disadvantage of stochastic algorithms is that they frequently do not find the optimal solution. Usually they find a suboptimal solution which is a local optimum of the evaluation function. In this thesis we describe the optimization of heterogeneous cargo transport between two locations with a group of vehicles. We begin by formally defining the problem in terms of cargo and vehicle characteristics. Then we present the evolutionary algorithm characteristics and their examples: genetic algorithms, evolutionary strategies, evolutionary programming, genetic programming and differential evolution. Next we describe the evolutionary algorithm implemented to solve the presented problem which was tested with the predefined parameter values on four test problems. The results were compared with those of the greedy algorithm. The evolutionary algorithm mostly found better results than the greedy algorithm. The algorithm parameters were tuned with a metaevolutionary algorithm which is also described. At the end we present the results obtained with the metaevolutionary algorithm, their comparison with the results of the greedy algorithm and the future work.
Secondary keywords:	cargo transport;group transport;transport optimization;evolutionary algorithm;parameter tuning;metaevolutionary algorithm;greedy algorithm;computer science;diploma;
File type:	application/pdf
Type (COBISS):	Undergraduate thesis
Thesis comment:	Univerza v Ljubljani, Fakulteta za računalništvo in informatiko
Pages:	IX, 46 str.
ID:	24260562