Rješavanje problema usmjeravanja vozila korištenjem neuronskih mreža

undergraduate thesis

• Title (english): Solving the vehicle routing problem using neural networks
• Author: Tin Jukić
• Mentor: Marko Đurasević (mentor)
• Committee member: Marko Đurasević (predsjednik povjerenstva)
• Committee member: Vjeko Kužina (član povjerenstva)
• Committee member: Nikolina Frid (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-07-15, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

U ovom radu se rješava problem usmjeravanja vozila korištenjem umjetnih neuronskih mreža. Opisuje se VRP problem (problem usmjeravanja vozila) te se pokušava riješiti jedna instanca iz Solomonove baze podataka. Korišteno je 25 automobila, svaki kapaciteta 200 te je bilo potrebno obići 100 korisnika. Zatim se opisuje jednostavni algoritam UCS, koji je naveden radi usporedbe s neuronskom mrežom, ali i radi objašnjavanja implementacije neuronske mreže za rješavanje ovog problema. Nakon toga slijedi opis genetskog algoritma, koji se koristi radi treniranja neuronske mreže. Zatim slijedi opis umjetne neuronske mreže te na kraju opis kako se genetski algoritam koristi za treniranje umjetne neuronske mreže. Slijedi opis konfiguracije korištene neuronske mreže. Na kraju se nalaze rezultati oba algoritma te usporedba dobivenih rezultata za ta dva algoritma. Rezultati su raznoliki te su doveli do zaključka da je u budućnosti potrebno koristiti bolje postavke za treniranje neuronske mreže.

Abstract (english)

Solving vehicle routing problem using artificial neural networks. Firstly, VRP problem is described and one instance of VRP from Solomon database is trying to be solved. There were 25 vehicles in the fleet, each with the same capacity of 200, and 100 customers were needed to be visited. Next, simple UCS algorithm was described, which was used for comparing VRP results with neural network results and to explain the implementation of the used neural network. Then, genetic algorithm was described, which is used for the trainign of neural network. After that, artificial neural network was described and then the genetic algorithm and previously described neural network were combined together to create an algorithm which was used for solving this instance of VRP problem. Next, configuration of the used neural network was described and after that given results from both algoritms were shown. At the end, the results were compared and the conclusion was made, which states that, in order to make neural network with genetic algorithm better, in the future, better settings for training the neural network are needed to be used.

• Keywords: VRP; Problem Usmjeravanja Vozila; Neuronska mreža; Neuronska mreža s genetskim algoritmom
• Keywords (english): VRP; Vehicle Routing Problem; Neural Network; Neural network with genetic algorithm
• Language: croatian
• URN:NBN: urn:nbn:hr:168:315149
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-15 10:25:41