2020_Performance Analysis Of SRM Family For Unconstrained Optimization Problems
| Format: | General Document |
|---|
| _version_ | 1860798167271866368 |
|---|---|
| building | INTELEK Repository |
| collection | Online Access |
| collectionurl | https://intelek.unisza.edu.my/intelek/pages/search.php?search=!collection3 |
| copyright | Copyright©PWB2025 |
| country | Malaysia |
| date | 2020-03-15 |
| format | General Document |
| id | 16263 |
| institution | UniSZA |
| originalfilename | PERFORMANCE ANALYSIS OF SRM FAMILY FOR UNCONSTRAINED OPTIMIZATION PROBLEMS.pdf |
| person | Syazni Binti Shoid |
| recordtype | oai_dc |
| resourceurl | https://intelek.unisza.edu.my/intelek/pages/view.php?ref=16263 |
| sourcemedia | Server storage Scanned document |
| spelling | 16263 https://intelek.unisza.edu.my/intelek/pages/view.php?ref=16263 https://intelek.unisza.edu.my/intelek/pages/search.php?search=!collection3 General Document Malaysia Library Staff (Top Management) Library Staff (Management) Library Staff (Support) Terengganu Faculty of Informatics & Computing English application/pdf 1.5 Server storage Scanned document Universiti Sultan Zainal Abidin UniSZA Private Access UNIVERSITI SULTAN ZAINAL ABIDIN SAMBox 2.4.24; modified using iTextSharp™ 5.5.10 ©2000-2016 iText Group NV (AGPL-version) Copyright©PWB2025 222 2020-03-15 PERFORMANCE ANALYSIS OF SRM FAMILY FOR UNCONSTRAINED OPTIMIZATION PROBLEMS.pdf Syazni Binti Shoid Switched Reluctance Motors Switched Reluctance Motors (SRM) Performance Analysis 2020_Performance Analysis Of SRM Family For Unconstrained Optimization Problems Conjugate gradient (CG) method is a well-known iterative method fbr solving large scale unconstrained optimization problems. This is attributable to the advantages of CG method such as low memory storage, simple algorithm and good convergence properties. Although many researchers have been done to improve this method, most of recent CG methods are complicated, difficult to implement and time consuming. This research proposes two new CG methods namely SRM and SRM2 based on the researcher's name (Syami, Rivaie and Mustafa). These methods are to improve the performance of CG methods especially in term of iteration numbers and Central Processing Unit (CPU) times. The SRM method is derived based on the combination of the Nur Hamizah, Rivaie and Mustafa (NRM) method and the Rivaie, Abashar, Mustafa and Ismail (RAMI) method. Meanwhile, the SRM2 method is a hybrid ol SRM and Hestenes and Stiet'el (HS) method. The proofs of convergence analysis for SRM and SRM2 under inexact line search are given. Then, the numerical comparisons are made by using three CG methods: the HS method, the Rivaie, Mustaf'a, Ismail and Leong (RMIL) method, and the RAMI method. They are tested with twenty-seven standard test functions, ranging from 2 to 5000 variables. For each test, four initial points of varying distance from the solution point are selected. All of the computational process is performed by MatlabR2Ol l software and the data generated are analyzed by using pertbrmance profile. This research shows that, the proposed methods satisfy the sufflcient descent condition and possess global convergence properties. Based on the pertbrmance profile, the performance shape shown by the number of iterations and CPU time are almost alike. This indicates that if the number iteration is increased, the CPU time will also increase. The SRM2 method is befter when compared to other methods which solve 100% of the test functions. Meanwhile, SRM and RAMI have solved 99.77%o and 99.09o/o of the functions, respectively. However, RMIL and HS can only solve the test functions below 90%. The results also show that HS and SRM2 are the lastest methods in solving the selected functions, followed by SRM, RAMI and RMIL methods. The SRM2 and SRM methods are considered to be efficient methods in solving the selected standard optimization functions with robust feature. Dissertations, Academic Unconstrained Optimization Thesis |
| spellingShingle | 2020_Performance Analysis Of SRM Family For Unconstrained Optimization Problems |
| state | Terengganu |
| subject | Switched Reluctance Motors Dissertations, Academic |
| summary | Conjugate gradient (CG) method is a well-known iterative method fbr solving large scale unconstrained optimization problems. This is attributable to the advantages of CG method such as low memory storage, simple algorithm and good convergence properties. Although many researchers have been done to improve this method, most of recent CG methods are complicated, difficult to implement and time consuming. This research proposes two new CG methods namely SRM and SRM2 based on the researcher's name (Syami, Rivaie and Mustafa). These methods are to improve the performance of CG methods especially in term of iteration numbers and Central Processing Unit (CPU) times. The SRM method is derived based on the combination of the Nur Hamizah, Rivaie and Mustafa (NRM) method and the Rivaie, Abashar, Mustafa and Ismail (RAMI) method. Meanwhile, the SRM2 method is a hybrid ol SRM and Hestenes and Stiet'el (HS) method. The proofs of convergence analysis for SRM and SRM2 under inexact line search are given. Then, the numerical comparisons are made by using three CG methods: the HS method, the Rivaie, Mustaf'a, Ismail and Leong (RMIL) method, and the RAMI method. They are tested with twenty-seven standard test functions, ranging from 2 to 5000 variables. For each test, four initial points of varying distance from the solution point are selected. All of the computational process is performed by MatlabR2Ol l software and the data generated are analyzed by using pertbrmance profile. This research shows that, the proposed methods satisfy the sufflcient descent condition and possess global convergence properties. Based on the pertbrmance profile, the performance shape shown by the number of iterations and CPU time are almost alike. This indicates that if the number iteration is increased, the CPU time will also increase. The SRM2 method is befter when compared to other methods which solve 100% of the test functions. Meanwhile, SRM and RAMI have solved 99.77%o and 99.09o/o of the functions, respectively. However, RMIL and HS can only solve the test functions below 90%. The results also show that HS and SRM2 are the lastest methods in solving the selected functions, followed by SRM, RAMI and RMIL methods. The SRM2 and SRM methods are considered to be efficient methods in solving the selected standard optimization functions with robust feature. |
| title | 2020_Performance Analysis Of SRM Family For Unconstrained Optimization Problems |
| title_full | 2020_Performance Analysis Of SRM Family For Unconstrained Optimization Problems |
| title_fullStr | 2020_Performance Analysis Of SRM Family For Unconstrained Optimization Problems |
| title_full_unstemmed | 2020_Performance Analysis Of SRM Family For Unconstrained Optimization Problems |
| title_short | 2020_Performance Analysis Of SRM Family For Unconstrained Optimization Problems |
| title_sort | 2020_performance analysis of srm family for unconstrained optimization problems |