一种超宽带与惯导融合的LSTM室内定位算法

针对复杂环境下的室内高精度定位需求，提出了一种超宽带和惯导融合定位方案。结合位置估计过程可被划分为时间序列预测问题的特点，提出了一种基于长短时记忆(Long Short Term Memory,LSTM) 网络的联合定位算法，并对其总体架构设计、数据预处理方法、网络结构设计、模型训练方法进行了研究。在此基础上，通过仿真和实测实验对联合定位算法进行验证，实验结果表明，该LSTM神经网络联合定位算法的定位精度优于传统TOA(Time of Arrival)、UKF(Unscented Kalman Filter)联合定位算法，适用复杂室内定位。

A Long Short Term Memory(LSTM) Indoor Positioning Algorithm Based on Fusion of UWB and Inertial Navigation

In order to improve the indoor positioning technology in complex environments,the position estimation process combining ultra-wideband(UWB) and inertial navigation positioning can be divided into the characteristics of time series prediction.The joint positioning algorithm of the network based on long short term memory(LSTM) is proposed,and its overall architecture design,data preprocessing method,network structure design and model training method are studied.On this basis,the joint localization algorithms are validated by simulation and actual experiments.Experimental results show that the positioning accuracy of the LSTM neural network joint positioning algorithm is superior to that of the traditional time of arrival(TOA) and unscented Kalman filter(UKF),and is suitable for complex indoor positioning.