文献类型：期刊文献

中文题名：基于策略学习的ERGA调度方法

英文题名：A Scheduling Method for Gas Accident Emergency Response Team Based on Strategy Learning

作者：胡超亭[1,2];冯翔[1,2];虞慧群[1,2]

机构：[1]华东理工大学信息科学与工程学院,上海200237;[2]上海智慧能源工程技术研究中心,上海200237

年份：2019

卷号：42

期号：4

起止页码：729

中文期刊名：山西大学学报（自然科学版）

外文期刊名：Journal of Shanxi University(Natural Science Edition)

收录：CSTPCD;;北大核心:【北大核心2017】；

基金：国家自然科学基金(61772200;61772201;61602175);上海市浦江人才计划(17PJ1401900);上海市经信委“信息化发展专项资金”(201602008)

语种：中文

中文关键词：ERGA跨区域调度;策略学习;资源调度问题;数据增维

外文关键词：ERGA cross-regional scheduling;strategy learning;resource scheduling problem;data enhancement

摘要：随着城市内部燃气管道逐渐增多,上海燃气事故愈加频繁,改进现有的燃气抢修调度方法显得尤为重要。通过研究上海燃气事故应急处置队伍(Emergency Response team for Gas Accidents,ERGA)跨区域调度问题,提出了一种基于策略学习的ERGA调度方法。首先构造以资源与任务的相关性为基础的关系模型,然后从区分时间、空间和特征3个方面进行数据增维操作,最后通过研究各种状态下的任务所具备的调度行为,形成多个"状态-行为"的映射关系。同时,将该调度方法应用到上海市ERGA跨区域调度问题中,从实验中验证了该方法的有效性,并与真实数据作对比,实验结果显示提升了6%的及时到场率,进一步表明了算法的可行性和优异性。
With the increasing amount of gas pipelines inside the city,gas accidents in Shanghai are becoming more frequent. So it is particularly important to improve the existing gas repair and dispatch methods. By studying the cross-regional scheduling problem for emergency response teams in gas accidents(ERGA) in Shanghai, a scheduling method for ERGA based on strategy learning is proposed. Firstly, the method constructs a relational model based on the correlation between resources and tasks. Then, data is dimensioned by distinguishing time, space and features. Finally, by studying the scheduling behavior of tasks in various states, a plurality of "state-behavior" mapping relationships are formed.Moreover,the scheduling method is applied to the ERGA cross-regional scheduling problem in Shanghai. The experimental results show that the 6% timely arrival rate is improved,which further indicates the feasibility and superiority of the algorithm.