Skalabilni model hidrauličke pumpe temeljen na umjetnoj neuronskoj mreži

undergraduate thesis

• Title (english): Scalable hydraulic pump model based on artificial neural network
• Author: Luka Ćavar
• Mentor: Mihael Cipek (mentor)
• Granter: University of Zagreb; Faculty of Mechanical Engineering and Naval Architecture; Zagreb
• Defense date and country: 2024-09-26, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Mechanical Engineering

Abstract

Ovaj rad bavi se razvojem skalabilnog modela neuronske mreže za predviđanje efikasnosti hidrauličke pumpe. Korištenjem MATLAB okruženja, cilj je bio kreirati model koji može efikasno obrađivati različite veličine i složenosti podataka bez gubitka performansi. Kroz analizu različitih arhitektura mreža, testirani su modeli na više skupova podataka kako bi se procijenila njihova sposobnost prilagođavanja. Četiri modela neuronske mreže su stvorena s obzirom na specifičnosti podataka i način pristupa s obzirom da su modeli trenirani na različitim dimenzijama pumpa kako bi se usporedila efikasnost i skalabilnost svakog modela, pri čemu je također dan uvid koliko je model neuronske mreže adekvatan za rješavanje problema predviđanja efikasnosti hidrauličke pumpe.

Abstract (english)

This study deals with the development of a scalable neural network model for predicting the efficiency of a hydraulic pump. Using the MATLAB environment, the aim was to create a model capable of efficiently processing different sizes and complexities of data without loss of performance. Through the analysis of various network architectures, the models were tested on multiple datasets to assess their adaptability. Four neural network models were created based on the specific data characteristics and approach, considering that the models were trained on different pump dimensions to compare the efficiency and scalability of each model. Additionally, insight was provided into how well the neural network model is suited for solving the problem of predicting hydraulic pump efficiency.

• Keywords: skalabilni model; neuronska mreža; MATLAB; predviđanje efikasnosti; hidraulička pumpa; arhitektura mreža; obrada podataka; skalabilnost; performanse modela.
• Keywords (english): scalable model; neural network; MATLAB; efficiency prediction; hydraulic pump; network architecture; data processing; scalability; model performance
• Language: croatian
• URN:NBN: urn:nbn:hr:235:191629
• Study programme: Title: Mechanical Engineering; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka strojarstva (sveučilišni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka strojarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2024-09-19 18:28:42