{"title":"用并行编程扫描输出空间求解源分析问题","authors":"Önder Halis Bettemi̇r, Tuğba Erzurum","doi":"10.18400/tekderg.595238","DOIUrl":null,"url":null,"abstract":"Exact Solution of Resource Leveling Problem by Exhaustive Enumeration with Parallel Programming Resource Leveling Problem (RLP) is solved by heuristic, meta-heuristic, and mathematical methods. However, the aforementioned methods cannot guarantee the exact solution for large size problems. In this study, number of feasible schedules which can be obtained by delaying the non-critical activities without violating the precedence relationships and elongating the project completion time are computed. All of the feasible schedules which can be defined as the search domain are enumerated and the guaranteed optimum solution for the RLP is obtained by a different method from the existing methods. Exponential equation between the search domain and the number of activities on serial path is derived and the insolvability of large RLP in a reasonable time by one central processing unit is verified. Partitioning of the problem into equal sizes is provided by parallel programming so that each particle contains the same number of enumeration. In this study, four RLP in which the largest problem has 36 activities are solved by exhaustive enumeration within reasonable solution time and it is proved that the proposed method is applicable. Exact solutions of larger problems can also be obtained by the proposed method if the problem is partitioned into smaller sizes.","PeriodicalId":49442,"journal":{"name":"Teknik Dergi","volume":null,"pages":null},"PeriodicalIF":0.8000,"publicationDate":"2021-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":"{\"title\":\"Kaynak Dengeleme Probleminin Arama Uzayını Paralel Programlama ile Tarayarak Kesin Çözümü\",\"authors\":\"Önder Halis Bettemi̇r, Tuğba Erzurum\",\"doi\":\"10.18400/tekderg.595238\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Exact Solution of Resource Leveling Problem by Exhaustive Enumeration with Parallel Programming Resource Leveling Problem (RLP) is solved by heuristic, meta-heuristic, and mathematical methods. However, the aforementioned methods cannot guarantee the exact solution for large size problems. In this study, number of feasible schedules which can be obtained by delaying the non-critical activities without violating the precedence relationships and elongating the project completion time are computed. All of the feasible schedules which can be defined as the search domain are enumerated and the guaranteed optimum solution for the RLP is obtained by a different method from the existing methods. Exponential equation between the search domain and the number of activities on serial path is derived and the insolvability of large RLP in a reasonable time by one central processing unit is verified. Partitioning of the problem into equal sizes is provided by parallel programming so that each particle contains the same number of enumeration. In this study, four RLP in which the largest problem has 36 activities are solved by exhaustive enumeration within reasonable solution time and it is proved that the proposed method is applicable. Exact solutions of larger problems can also be obtained by the proposed method if the problem is partitioned into smaller sizes.\",\"PeriodicalId\":49442,\"journal\":{\"name\":\"Teknik Dergi\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.8000,\"publicationDate\":\"2021-05-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"3\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Teknik Dergi\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://doi.org/10.18400/tekderg.595238\",\"RegionNum\":4,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"ENGINEERING, CIVIL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Teknik Dergi","FirstCategoryId":"5","ListUrlMain":"https://doi.org/10.18400/tekderg.595238","RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"ENGINEERING, CIVIL","Score":null,"Total":0}
Kaynak Dengeleme Probleminin Arama Uzayını Paralel Programlama ile Tarayarak Kesin Çözümü
Exact Solution of Resource Leveling Problem by Exhaustive Enumeration with Parallel Programming Resource Leveling Problem (RLP) is solved by heuristic, meta-heuristic, and mathematical methods. However, the aforementioned methods cannot guarantee the exact solution for large size problems. In this study, number of feasible schedules which can be obtained by delaying the non-critical activities without violating the precedence relationships and elongating the project completion time are computed. All of the feasible schedules which can be defined as the search domain are enumerated and the guaranteed optimum solution for the RLP is obtained by a different method from the existing methods. Exponential equation between the search domain and the number of activities on serial path is derived and the insolvability of large RLP in a reasonable time by one central processing unit is verified. Partitioning of the problem into equal sizes is provided by parallel programming so that each particle contains the same number of enumeration. In this study, four RLP in which the largest problem has 36 activities are solved by exhaustive enumeration within reasonable solution time and it is proved that the proposed method is applicable. Exact solutions of larger problems can also be obtained by the proposed method if the problem is partitioned into smaller sizes.
期刊介绍:
The scope of Teknik Dergi is naturally confined with the subjects falling in the area of civil engineering. However, the area of civil engineering has recently been significantly enlarged, even the definition of civil engineering has somewhat changed.
Half a century ago, engineering was simply defined as “the art of using and converting the natural resources for the benefit of the mankind”. Today, the same objective is expected to be realised (i) by complying with the desire and expectations of the people concerned and (ii) without wasting the resources and within the sustainability principles. This change has required an interaction between engineering and social and administrative sciences. Some subjects at the borderline between civil engineering and social and administrative sciences have consequently been included in the area of civil engineering.
Teknik Dergi defines its scope in line with this understanding. However, it requires the papers falling in the borderline to have a significant component of civil engineering.