Yufan Zhou, Xinhua Yang, Ailing Shen, Juan Lin, Yiwen Zhong
{"title":"Simulated Annealing Algorithm Based on Single-direction Greedy Decoding for Solving Corridor Allocation Problem","authors":"Yufan Zhou, Xinhua Yang, Ailing Shen, Juan Lin, Yiwen Zhong","doi":"10.1145/3529836.3529848","DOIUrl":null,"url":null,"abstract":"The Corridor Allocation Problem (CAP) is an NP-hard combinatorial optimization problem which aims to find the optimal layout of facilities on both side of a corridor, so as to minimize the flow cost between all pairs of facilities. Most existing metaheuristics use permutation of facilities to represent a solution, then a decoding strategy is used to map the solution representation into a layout. The decoding strategies used by those metaheuristics may lead to inconsistence between solution representation and the corresponding layout. For example, two facilities, which are far apart from each other in the representation, may become adjacent to each other in the layout. This inconsistence may affect the performance of metaheuristic. To overcome this shortage, this paper presents a Single-direction Greedy Decoding (SGD) strategy to map a permutation-based solution representation into a layout. Using the SGD strategy, a Hybrid Simulated Annealing (HSA) is proposed for solving the CAP. In HSA, a hybrid neighborhood structure is designed to produce candidate solutions. The HSA algorithm is experimentally analyzed on 23 benchmark instances with up to 70 facilities. Experimental results confirm the advantage of the SGD strategy and the hybrid neighborhood structure. Furthermore, the HSA algorithm found new optimal solutions on 13 instances.","PeriodicalId":285191,"journal":{"name":"2022 14th International Conference on Machine Learning and Computing (ICMLC)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 14th International Conference on Machine Learning and Computing (ICMLC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3529836.3529848","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
The Corridor Allocation Problem (CAP) is an NP-hard combinatorial optimization problem which aims to find the optimal layout of facilities on both side of a corridor, so as to minimize the flow cost between all pairs of facilities. Most existing metaheuristics use permutation of facilities to represent a solution, then a decoding strategy is used to map the solution representation into a layout. The decoding strategies used by those metaheuristics may lead to inconsistence between solution representation and the corresponding layout. For example, two facilities, which are far apart from each other in the representation, may become adjacent to each other in the layout. This inconsistence may affect the performance of metaheuristic. To overcome this shortage, this paper presents a Single-direction Greedy Decoding (SGD) strategy to map a permutation-based solution representation into a layout. Using the SGD strategy, a Hybrid Simulated Annealing (HSA) is proposed for solving the CAP. In HSA, a hybrid neighborhood structure is designed to produce candidate solutions. The HSA algorithm is experimentally analyzed on 23 benchmark instances with up to 70 facilities. Experimental results confirm the advantage of the SGD strategy and the hybrid neighborhood structure. Furthermore, the HSA algorithm found new optimal solutions on 13 instances.