摘要
针对DV-Hop定位模型跳距估计精度低且难准确获取未知节点位置的问题,提出改进型DV-Hop定位优化模型及其求解的混合视觉进化神经网络优化算法。DV-Hop定位模型设计中,引入4种通信半径细化节点间的跳数计算,并利用动态权重因子修正最小跳数及平均跳距计算模型,获得改进型DV-Hop的计算模型。算法设计中,依据果蝇和蝗虫视觉系统的信息处理机制,建立输出全局和局部学习率的混合视觉神经网络,进而在此神经网络的输出引导下,借助哈里斯鹰优化算法中哈里斯鹰的位置更新策略设计状态更新策略,由此获得能求解DV-Hop定位问题的视觉进化神经网络优化算法。比较性的数值实验显示,该算法求解基准函数优化问题具有明显优势,并对未知节点的定位精度高且收敛速度快。
This work develops an improved DV-Hop optimization model and the related optimization approach,in order to solve the problems of the low accuracy of jump distance estimation in the DV-Hop positioning algorithm and the difficulty of accurately obtaining the positions of unknown nodes.In the design of the model,four kinds of communication radii are introduced to calculate the hop numbers between nodes,and later,the original minimum hop number and average hop distance models are improved in terms of dynamic weight factors.In the design of the algorithm,based on the visual information-processing mechanism of the fruit fly and locust visual systems,a hybrid visual neural network,which can generate global and local learning rates,is developed and integrated with a Harris Hawk optimization-based state update strategy to construct a hybrid visual neural network optimization approach.The comparative experiments have validated that,when solving the benchmark function optimization problems,the algorithm is of significant advantage over the compared approaches and can locate the unknown nodes with high accuracy and rapid convergence.
作者
郑悦
张著洪
ZHENG Yue;ZHANG Zhuhong(College of Big Data and Information Engineering,Guizhou University,Guiyang 550025,China;Guizhou Provincial Characteristic Key Laboratory of System Optimization and Scientific Computation,Guizhou University,Guiyang 550025,China)
出处
《智能计算机与应用》
2023年第3期1-9,15,共10页
Intelligent Computer and Applications
基金
国家自然科学基金(62063002)。
