[1]徐静如,董红斌,赵炳旭,等.具有角色意识的社区服务型时空众包任务分配[J].智能系统学报,2023,18(2):293-304.[doi:10.11992/tis.202205017]
XU Jingru,DONG Hongbin,ZHAO Bingxu,et al.Community service-oriented spatiotemporal crowdsourcing task allocation with role awareness[J].CAAI Transactions on Intelligent Systems,2023,18(2):293-304.[doi:10.11992/tis.202205017]
点击复制
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
18
期数:
2023年第2期
页码:
293-304
栏目:
学术论文—智能系统
出版日期:
2023-05-05
- Title:
-
Community service-oriented spatiotemporal crowdsourcing task allocation with role awareness
- 作者:
-
徐静如, 董红斌, 赵炳旭, 冀若含
-
哈尔滨工程大学 计算机科学与技术学院,黑龙江 哈尔滨 150001
- Author(s):
-
XU Jingru, DONG Hongbin, ZHAO Bingxu, JI Ruohan
-
College of Computer Science and Technology, Harbin Engineering University, Harbin 150001, China
-
- 关键词:
-
社区服务型时空众包; 任务分配; E-CARGO; 基于角色协作; 核密度聚类; 角色感知; 学习遗忘曲线; 地点资格值多阶段量化
- Keywords:
-
community service-oriented spatiotemporal crowdsourcing; task allocation; E-CARGO; role-based collaboration; kernel density clustering; role perception; learning and forgetting curve; multistage quantification of site qualification value
- 分类号:
-
TP311
- DOI:
-
10.11992/tis.202205017
- 摘要:
-
为了适应社区众包配送中投递箱或代收点收纳量有限等情况,本文提出社区服务型时空众包任务分配问题,根据用户定义的时间将物品较为实时地配送到地。针对该问题,本文通过基于角色的协作模型E-CARGO(environments-classes, agents, roles, groups, objects)形式化问题,针对高资格值工人配送高价值量订单集的目标,提出基于贪婪分配的PQGR(places-qualification-based greedy)算法、基于考虑代理和角色冲突的团队多角色分配方法的PQGM(places-qualification-based GMAC)算法以及进一步缩短运行时间的改进PQGM算法。数据处理和量化方面,提出基于核密度聚类的新型角色感知方法以实现任务的有效划分,提出基于学习遗忘曲线的代理地点资格值多阶段量化模型,实现代理地点资格值的在线学习和自适应更新。最后,本文在gMission数据集和合成数据集上进行实验,验证了算法的有效性和效率。
- Abstract:
-
This paper proposes the community service-oriented spatiotemporal crowdsourcing task allocation problem to adapt to the limited amount of delivery boxes or collection points in community crowdsourcing. The articles will be relatively delivered to the ground in real time according to the user-defined time. This paper proposes a PQGR algorithm based on greedy allocation, a PQGM algorithm based on the GMRACRA method, and a PQGMA algorithm that further shortens the running time to address the aforementioned problem. This algorithm aims to realize the objective of delivering high-value order sets by workers with high qualification value through the formalization problem of E-CARGO based on the role cooperation model. Considering data processing and quantification, a new role perception method based on kernel density clustering is proposed to realize effective task division. The places-qualification-based multistage quantitative model of an agent is proposed on the basis of learning and forgetting curves to realize online learning and adaptive updating of agent location qualification value. Finally, experiments on gMission and synthetic datasets are conducted to verify the effectiveness and efficiency of the algorithm.
备注/Memo
收稿日期:2022-05-16。
基金项目:黑龙江自然科学基金项目(LH2020F023).
作者简介:徐静如,硕士研究生,主要研究方向为时空众包、群体智能;董红斌,教授,博士生导师,中国计算机学会高级会员,主要研究方向为机器学习、人工智能、多智能体系统和数据挖掘。主持和完成国家自然科学基金、工信部基础研究项目、黑龙江省自然科学基金项目,荣获黑龙江省高校科学技术奖和黑龙江省优秀高等教育科学成果奖。主编教材2部。发表学术论文90余篇;赵炳旭,博士,主要研究方向为时空众包、移动计算和人工智能
通讯作者:董红斌.E-mail: donghongbin@hrbeu.edu.cn
更新日期/Last Update:
1900-01-01