| 摘要: |
| 针对采用遗传算法(Genetic Algorithm,GA)优化广义回归神经网络(Generalized Regression Neural Network,GRNN)的射频识别（Radio Frequency Identification，RFID）室内定位模型存在的早熟、收敛速度慢、不能保证解是全局最优等问题，提出采用思维进化算法(Mind Evolutionary Algorithm,MEA)来寻找广义神经网络的最优光滑因子，从而确定最优定位模型。首先用GRNN建立节点定位模型，阅读器与标签的接收信号强度值作为GRNN的输入，节点坐标作为输出，根据适应度函数值，通过MEA寻找GRNN的最优平滑参数。实验结果表明，通过MEA优化的GRNN模型的定位精度比GA优化的GRNN定位模型的精度高、泛化能力强，并且比后者的效率高,能够避免GA陷入局部最优的问题。 |
| 关键词: 室内定位 射频识别 广义神经网络 思维进化算法优化 |
| DOI： |
|
| 基金项目:国家自然科学基金资助项目(61340005)；北京市自然科学基金面上项目(4132012) |
|
| RFID Indoor Positioning Based on Mind Evolutionary Algorithm Optimization GRNN |
| SONG Ningjia,CUI Yinghua |
| (School of Information and Communication Engineering，Beijing Information Science and Technology University，Beijing 100101,China) |
| Abstract: |
| The radio frequency identification(RFID) indoor positioning model based on genetic algorithm(GA)-generalized regression neural network(GRNN) has such problems as prematurity,low convergence speed and non-optimal solution.To solve above problems,the mind evolutionary algorithm(MEA) is used to find the optimal smoothing factor of the generalized neural network to determine the optimal positioning model.Firstly,the GRNN is used to establish the node localization model.The received signal strength value of the reader and the tag is used as the input of the GRNN,and the node coordinates are used as the output.According to the fitness function value,the excellent smoothing parameters of GRNN are found by MEA.The experimental results show that the positioning accuracy of the GRNN model optimized by MEA is higher than that of the GA-optimized GRNN positioning model,and the generalization ability is strong,and the efficiency is higher,which can avoid the problem that GA falls into local optimum. |
| Key words: indoor positioning RFID generalized neural network mind evolutionary algorithm optimization |
《电讯技术》编辑部 