基于Bi-LSTM神经网络的室内可见光定位方法OA
Indoor visible light positioning method based on Bi-LSTM neural network
双向长短时记忆(Bi-LSTM)神经网络由于超参数众多,难以获得最优系统模型.同时,考虑到灰狼优化(GWO)算法可能过早收敛的情况,提出了一种采用GWO结合粒子群(GWO-PSO)算法优化Bi-LSTM神经网络的单灯定位方法.通过优化网络中的学习率、隐藏神经元个数等超参数,提高系统的稳定性和定位精度.最后,采用加权K邻近(WKNN)算法对误差较大的点进行优化,以获得更精确的定位位置.仿真结果表明,在 3 m×3.6 m×3 m的室内环境中,所提定位方法的平均定位误差为 3.57 cm,其中90%的定位误差在 6cm内.
Due to the large number of hyperparameters,it is difficult to obtain the optimal system model for the bidirectional long short-term memory(Bi-LSTM)neural network.At the same time,considering the possibility of premature convergence in the Grey Wolf optimizer(GWO)algorithm,a single-lamp localization method using the GWO combined with particle swarm opti-mization(GWO-PSO)algorithm to optimize the Bi-LSTM neural network is proposed.By optimizing hyperparameters such as the learning rate and the number of hidden neurons in the network,the stability and positioning accuracy of the system are im-proved.Finally,the weighted K-nearest neighbors(WKNN)algorithm is used to optimize points with large errors to obtain more accurate positioning locations.The simulation results show that in an indoor environment of 3 m×3.6 m×3 m,the average posi-tioning error of the proposed localization method is 3.57 cm,with 90%of the positioning errors within 6 cm.
王乐乐;秦岭;胡晓莉;赵德胜
内蒙古科技大学信息工程学院,内蒙古包头 014010
电子信息工程
可见光定位双向长短时记忆灰狼结合粒子群加权K近邻
visible light positioningbidirectional long short term memoryGrey Wolf combined with particle swarm optimiza-tion algorithmweighted K-nearest neighbor
《光通信技术》 2024 (002)
36-41 / 6
国家自然科学基金项目(62161041)资助;内蒙古自治区应用技术研究与开发资金项目(2021GG0104)资助.
评论