基于信任度的合一系统社会任务分配

发布时间：2018-06-23 04:21

本文选题：社会网络 + 任务分配　；参考：《软件学报》2017年07期

【摘要】：在多智体社会网络中,传统的任务分配模型一般采用直接面向任务执行者的分配机制.它们不考虑社会网络组织结构对任务分配性能的巨大影响,也很少透彻地研究不可靠社会中的任务分配.针对这些问题,开创性地研究了软/硬件合一系统的任务分配,即按递阶、分层的思想设计了协作组织模型,并基于此提出了面向社区基于社会协调"软件人"的任务分配模型.模型研究过程中,提出了基于直接信任度和社区声誉的社区信任度评估机制、基于社区信任度和社区物理能力的节点选择机制、基于负载均衡的社区内任务分配机制和基于上下文资源的任务再分配策略.实验结果表明:与常见的直接面向任务执行者和基于资源的任务分配模型相比,所提出的模型具有更优的任务分配性能,且对社会任务环境变化具有更好的鲁棒性;社区内基于负载均衡的分配机制和基于上下文资源的再分配策略也有效提高了分配性能,降低了网络中的通信密度.
[Abstract]:In the multi intelligence social network, the traditional task allocation model generally adopts the allocation mechanism of direct task oriented implementers. They do not take into account the great influence of the social network organization structure on the task allocation performance, and rarely study the task allocation in the unreliable society. The task allocation of a system is designed according to the hierarchical and hierarchical idea, and based on this, a task allocation model for community based social coordination is proposed. In the process of the model study, a community trust evaluation mechanism based on direct trust and community reputation is proposed, based on community trust and community physics. The node selection mechanism of the capability is based on the load balancing task allocation mechanism in the community and the task redistribution strategy based on the context resource. The experimental results show that compared with the common task oriented and resource based task allocation model, the proposed model has a better task allocation performance and a social task ring. The environment changes have better robustness, and the distribution mechanism based on load balancing and the redistribution strategy based on context resources in the community can also effectively improve the distribution performance and reduce the communication density in the network.
【作者单位】：辽宁工程技术大学软件学院;北京科技大学计算机与通信工程学院;
【基金】：国家自然科学基金(61404069,61401185,61540056) 辽宁省教育厅科学技术项目(LJYL052)~~
【分类号】：TP18;TP242

【相似文献】