Faster Capacitated Arc Routing: A Sequence-to-Sequence Approach

被引:2
作者
Hong, Wenjing [1 ,2 ]
Liu, Tonglin [3 ]
机构
[1] Southern Univ Sci & Technol, Guangdong Prov Key Lab Brain Inspired Intelligent, Dept Comp Sci & Engn, Shenzhen 518055, Peoples R China
[2] Guangdong Hong Kong Macao Greater Bay Area Ctr Br, Guangzhou 510515, Peoples R China
[3] Beijing Electromech Engn Inst, Sci & Technol Complex Syst Control & Intelligent, Beijing 100074, Peoples R China
基金
中国国家自然科学基金;
关键词
Task analysis; Heuristic algorithms; Training; Optimization; Deep learning; Routing; Machine learning algorithms; Capacitated arc routing problem; sequence-to-sequence; deep neural network; NEURAL-NETWORK; ALGORITHMS; OPTIMIZATION;
D O I
10.1109/ACCESS.2022.3140783
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The Capacitated Arc Routing Problem (CARP) is an NP-hard optimization problem that has been investigated for decades. Heuristic search methods are commonly used to solve it. However, given a CARP instance, most heuristic search algorithms require plenty of time to iteratively search for the solution from scratch, and hence may be impractical for emerging applications that need a solution to be obtained in a very short time period. In this work, a novel approach to efficiently solve CARP is presented. The proposed approach replaces the heuristic search process with the inference phase of a trained Deep Neural Network (DNN), which is trained to take a CARP instance as the input and outputs a solution to the instance. In this way, CARP could be solved by a direct mapping rather than by iterative search, and hence could be more efficient and more easily accelerated by the use of GPUs. Empirical study shows that the DNN-based solver can achieve significant speed-up with minor performance loss, and up to hundreds of times acceleration in extreme cases.
引用
收藏
页码:4777 / 4785
页数:9
相关论文
共 43 条
[1]   A Sparse Branch and Bound Optimization of Noisy Weighted DAG Modification Under Constraints: A Method for Monocular Data Association to Multiple Laser Planes [J].
Al-Osaimi, Faisal R. .
IEEE ACCESS, 2021, 9 :5994-6007
[2]   Accelerating compute intensive medical imaging segmentation algorithms using hybrid CPU-GPU implementations [J].
Alsmirat, Mohammad A. ;
Jararweh, Yaser ;
Al-Ayyoub, Mahmoud ;
Shehab, Mohammed A. ;
Gupta, Brij B. .
MULTIMEDIA TOOLS AND APPLICATIONS, 2017, 76 (03) :3537-3555
[3]  
[Anonymous], 2017, P INT C LEARN REPR T
[4]   An efficiency-based path-scanning heuristic for the capacitated arc routing problem [J].
Arakaki, Rafael Kendy ;
Usberti, Fabio Luiz .
COMPUTERS & OPERATIONS RESEARCH, 2019, 103 :288-295
[5]  
Bahdanau D, 2016, Arxiv, DOI arXiv:1409.0473
[6]  
Bellemare MG, 2017, PR MACH LEARN RES, V70
[7]   A review on estimation of distribution algorithms in permutation-based combinatorial optimization problems [J].
Ceberio, Josu ;
Irurozki, Ekhine ;
Mendiburu, Alexander ;
Lozano, Jose A. .
PROGRESS IN ARTIFICIAL INTELLIGENCE, 2012, 1 (01) :103-117
[8]   A hybrid metaheuristic approach for the capacitated arc routing problem [J].
Chen, Yuning ;
Hao, Jin-Kao ;
Glover, Fred .
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 2016, 253 (01) :25-39
[9]  
Dai HJ, 2017, ADV NEUR IN, V30
[10]  
Dai HJ, 2016, PR MACH LEARN RES, V48