基于I-Greedy求解CVaR模型的传感器网络布局优化  

Optimization of Sensor Network Layout Based on I-Greedy Solution CVaR Model

作　　者：高安迪 吴晓霞[1] 李峰 GAO An-di;WU Xiao-xia;LI Feng(Department of Information Engineering,Hebei Vocational College of Mechanical and Electrical Technology,Hebei Xingtai 054000,China;College of Information Science and Engineering,Hebei University of Science and Technology,Hebei Shijiazhuang 050018,China)

机构地区：[1]河北机电职业技术学院,信息工程系,河北邢台054000 [2]河北科技大学,信息科学与工程学院,河北石家庄050018

出　　处：《机械设计与制造》2023年第12期138-141,共4页Machinery Design & Manufacture

基　　金：河北省高等学校科学研究项目(SQ2021126)。

摘　　要：无线传感器网络在机械设备状态监测领域有着重要的作用,为了解决传统随机优化方法不适合某些复杂场传感器网络布局景的问题,提出了一种基于I-Greedy求解CVaR模型。采用惰性赋值的方法完成算法的简化过程,为τ设置了相应的搜索间隔Δ和搜索区间(0,Γ),防止算法出现局部最优解的情况,通过惰性赋值的方式实现快速搜索的功能。研究结果表明:τ搜索上界Γ设定成50和置信水平α=0.9时,能够确保各置信水平都搜索获得全局最优解。逐渐增加传感器节点数量后,布局效益获得了持续提升。计算得到互信息相对随机部署方法增加69%,与传统贪婪算法相比增加14.1%。CVaR布局模型相对传统布局模型可以达到更低损失程度,能够获得更优布局结果,提升了模型鲁棒性。算法能够显著降低时间复杂度,特别是进行大规模传感器布局时表现出了更强的优越性。In order to solve the problem that traditional stochastic optimization method is not suitable for some complex field sensor network layout,an I-Greedy solution CVaR model is proposed.The method of lazy assignment is used to complete the simplification of the algorithm,and the corresponding search intervalΔand search interval(0,Γ)are set forτto prevent the occurrence of local optimal solution in the algorithm.You can realize the function of fast search by means of lazy assignment.The results show that whenΓofτsearch upper bound is set to 50 and confidence levelα=0.9,the global optimal solution can be obtained by searching each confidence level.After increasing the number of sensor nodes gradually,the layout benefit is improved continuously.The calculated mutual information increased by 69%compared with the random deployment method and 14.1%compared with the traditional greedy algorithm.Compared with the traditional layout model,CVaR layout model can achieve a lower loss degree,obtain better layout results,and improve the robustness of the model.The algorithm can significantly reduce the time complexity,especially for large-scale sensor layout.

关 键 词：传感器网络 布局优化 贪婪算法 布局损失 

分 类 号：TH16[机械工程—机械制造及自动化] TN92[电子电信—通信与信息系统]