Pregled optimizacijskih metod in optimizacija sinhronskega motorja s trajnimi magneti : magistrsko delo

Stare, Živa (author), Miljavec, Damijan (mentor)

V magistrski nalogi so opisane optimizacijske metode, ki so del računalniškega programa Matlab, ter predstavljena optimizacija z roji delcev na električnem sinhronskem motorju s trajnimi magneti. Magistrska naloga se deli na dva večja sklopa. V prvem sklopu so opisane vse optimizacijske metode, ki so na voljo v Matlab globalno optimizacijskem programskem paketu. Poleg popularnih naravnih algoritmov, kot so genetski algoritmi, simulirano žarjenje in optimizacija z roji delcev, so v paketu dodane tudi manj znane, a še vedno pogosto uporabljene metode, kot so optimizacija z iskanjem vzorcev, optimizacija z nadomestnimi deli ter multi-start optimizacija in globalno iskanje. Programski paket omogoča tudi večkriterijsko optimizacijo. Pregled optimizacijskih metod je razdeljen na sam opis metode, potek optimizacije in predstavitev prednosti in slabosti omenjene metode. Drugi sklop je namenjen preizkušanju metode z roji delcev na problemu optimizacije sinhronskega motorja s površinsko nameščenimi trajnimi magneti. Pred samo optimizacijo je opisan tudi potek načrtovanja sinhronskega motorja na podlagi podanih zahtevanih nazivnih podatkov. Analitično načrtovan sinhronski motor je bil sprva optimiziran na podlagi kriterija, da ima motor čim višjo vrednost kvocienta nazivnega navora in mase železnih delov motorja. Optimizirani motor ima vrednost kvocienta za 23,4-% višjo kot prvotni motor. Nato je bil isti prvotni sinhronski motor še optimiziran na podlagi kriterija, da ima motor čim višjo vrednost produkta nazivnega navora in izkoristka motorja. Optimizirani motor ima vrednost produkta za 76,2-% višjo kot prvotni motor. Na vseh treh verzijah sinhronskega motorja so bile izvedene tudi simulacije v simulacijskem programu FEMM 4.2. Simulirani so bili poteki gostote magnetnega pretoka v prerezu motorja ter navorni karakteristiki nazivnega in samodržnega navora. Pri optimiziranih verzijah motorja se je poleg povišanja amplitudne vrednosti nazivnega navora povečala tudi amplitudna vrednost samodržnega navora.

• Language: Slovenian
• Keywords: optimizacija, optimizacija z roji delcev, Matlab, sinhronski motor s trajnimi magneti
• Work type: Master's thesis/paper
• Typology: 2.09 - Master's Thesis
• Organization: FE - Faculty of Electrical Engineering
• Place of publishing: Ljubljana
• Publisher: [Ž. Stare]
• Year: 2021
• Number of pages: 134 str.
• PID: 20.500.12556/RUL-130136
• UDC: 621.313.323(043.3)
• COBISS.SI-ID: 76142595
• Publication date in RUL: 10.09.2021

Secondary language

• Language: English
• Title: Overview of optimization methods and optimization of permanent magnet synchronous motor : magistrski študijski program druge stopnje Elektrotehnika

Abstract

The master's thesis describes optimization methods that are part of the Matlab computer program and presents particle swarm optimization performed on an electric permanent magnet synchronous motor. The master's thesis is divided into two major parts. The first part describes all the optimization methods available in the Matlab global optimization software package. In addition to popular natural algorithms such as genetic algorithms, simulated annealing and particle swarm optimization, the package also includes less known but still commonly used methods such as pattern search optimization, surrogate optimization, multi-start optimization and global search. The software package also enables multiobjective optimization. The overview of the optimization methods is divided into a description of the methods, the course of an optimization and a presentation of the advantages and disadvantages of this method. The second part is intended for testing the particle swarm method on the problem of optimizing a synchronous motor with surface-mounted permanent magnets. Prior to the optimization, the design process of the synchronous motor is described on the basis of the given required nominal data. The analytically designed synchronous motor was initially optimized based on the criteria that the motor has the highest possible value of the quotient of the nominal torque and the mass of the iron parts of the motor. The optimized engine has a quotient value 23,4-% higher than the original engine. Then, the same original synchronous motor was further optimized based on the criteria that the motor has the highest possible value of the product of nominal torque and motor efficiency. The optimized motor has a product value 76,2-% higher than the original one. Calculations using the FEMM 4.2 simulation program were also performed on all three versions of the synchronous motor. The magnetic flux density distribution in the cross section of the motor was calculated, as well as the nominal torque characteristics and cogging torque. In the optimized versions, in addition to increasing the amplitude value of the nominal torque, the amplitude value of the cogging torque was also increased.

• Keywords: optimization, particle swarm optimization, Matlab, permanent magnet synchronous motor